From 732cea1847dbd4e91cd21fce23ebda06bb9cface Mon Sep 17 00:00:00 2001 From: Thurgood Date: Tue, 26 May 2026 22:24:46 +0000 Subject: [PATCH] Rollback legal skills catalog to 883 entries --- .../SKILL.md | 33 --------- .../alphalaw-medical-chronology/SKILL.md | 32 --------- .../legal/alphalaw-pi-case-workflow/SKILL.md | 31 --------- skills/legal/anytime-agent-workflow/SKILL.md | 47 ------------- skills/legal/anytime-demand-letter/SKILL.md | 46 ------------- .../legal/anytime-medical-chronology/SKILL.md | 49 ------------- .../legal/automated-contract-summary/SKILL.md | 38 ---------- .../botmediation-adr-settlement/SKILL.md | 33 --------- .../SKILL.md | 33 --------- skills/legal/cardamon-gap-analysis/SKILL.md | 56 --------------- .../legal/cardamon-horizon-scanning/SKILL.md | 53 -------------- .../cardamon-obligation-mapping/SKILL.md | 46 ------------- skills/legal/case-law-research/SKILL.md | 29 -------- .../SKILL.md | 37 ---------- .../codex-claims-rebuttal-drafting/SKILL.md | 37 ---------- .../SKILL.md | 32 --------- skills/legal/compli-gap-assessment/SKILL.md | 38 ---------- .../compliance-framework-mapping/SKILL.md | 38 ---------- .../construction-code-compliance/SKILL.md | 35 ---------- .../SKILL.md | 36 ---------- .../construction-contract-review/SKILL.md | 35 ---------- .../SKILL.md | 36 ---------- .../construction-defect-analysis/SKILL.md | 31 --------- .../construction-drawing-review/SKILL.md | 32 --------- .../legal/contract-leakage-analysis/SKILL.md | 29 -------- .../legal/contract-renewal-tracking/SKILL.md | 29 -------- skills/legal/contract-review-triage/SKILL.md | 38 ---------- .../legal/contract-term-extraction/SKILL.md | 28 -------- .../covenant-fund-investment-review/SKILL.md | 37 ---------- .../covenant-mfns-election-analysis/SKILL.md | 37 ---------- .../legal/covenant-nda-markup-triage/SKILL.md | 36 ---------- .../legal/deepjudge-agile-workflows/SKILL.md | 24 ------- .../legal/deepjudge-knowledge-search/SKILL.md | 24 ------- .../SKILL.md | 24 ------- .../legal/deposition-interrogation/SKILL.md | 35 ---------- .../deposition-searchable-record/SKILL.md | 35 ---------- .../legal/deposition-video-analysis/SKILL.md | 35 ---------- .../SKILL.md | 37 ---------- .../SKILL.md | 37 ---------- .../dexterity-version-control-deals/SKILL.md | 37 ---------- .../dyspute-ai-mediation-assistant/SKILL.md | 35 ---------- .../SKILL.md | 35 ---------- .../ecommerce-infringement-analysis/SKILL.md | 38 ---------- .../enforce-autonomous-takedown/SKILL.md | 38 ---------- skills/legal/expert-witness-matching/SKILL.md | 31 --------- .../legal/firstdrafts-case-summary/SKILL.md | 34 --------- .../SKILL.md | 36 ---------- .../firstdrafts-litigation-motion/SKILL.md | 35 ---------- .../SKILL.md | 34 --------- .../flipthrough-negotiation-briefing/SKILL.md | 35 ---------- .../SKILL.md | 35 ---------- .../genesis-case-evaluation-drafting/SKILL.md | 35 ---------- .../legal/genesis-medical-chronology/SKILL.md | 35 ---------- .../legal/genesis-tort-case-summary/SKILL.md | 38 ---------- skills/legal/habeas-case-strategy/SKILL.md | 49 ------------- .../legal/habeas-citation-synthesis/SKILL.md | 47 ------------- skills/legal/habeas-custom-agent/SKILL.md | 48 ------------- skills/legal/haloo-brand-monitoring/SKILL.md | 24 ------- .../legal/haloo-portfolio-management/SKILL.md | 24 ------- .../legal/haloo-trademark-clearance/SKILL.md | 24 ------- skills/legal/ip-portfolio-monitoring/SKILL.md | 39 ----------- .../lawxy-case-sentiment-analysis/SKILL.md | 34 --------- .../lawxy-contract-review-studio/SKILL.md | 34 --------- skills/legal/lawxy-due-diligence/SKILL.md | 34 --------- .../legal/lawxy-legal-agent-network/SKILL.md | 35 ---------- .../legal/lawy-contract-intelligence/SKILL.md | 60 ---------------- .../legal/lawy-matter-summarization/SKILL.md | 58 ---------------- .../lawy-verified-legal-research/SKILL.md | 54 --------------- skills/legal/legal-translation/SKILL.md | 29 -------- skills/legal/legal1up-discovery/SKILL.md | 48 ------------- skills/legal/legal1up-joint-appendix/SKILL.md | 48 ------------- skills/legal/litigation-drafting/SKILL.md | 29 -------- skills/legal/med-mal-case-evaluation/SKILL.md | 30 -------- .../ml-contract-provision-extraction/SKILL.md | 38 ---------- .../legal/pactum-negotiation-agent/SKILL.md | 50 -------------- skills/legal/pactum-price-list/SKILL.md | 48 ------------- .../patent-infringement-claim-chart/SKILL.md | 35 ---------- .../patent-infringement-detection/SKILL.md | 35 ---------- .../patent-invention-disclosure/SKILL.md | 37 ---------- .../legal/patent-portfolio-analysis/SKILL.md | 35 ---------- .../legal/patentability-assessment/SKILL.md | 32 --------- .../legal/patentprior-art-analysis/SKILL.md | 35 ---------- .../legal/paxton-deposition-outline/SKILL.md | 34 --------- .../legal/paxton-legal-research-memo/SKILL.md | 33 --------- .../legal/paxton-policy-issue-memo/SKILL.md | 33 --------- .../permitting-diligence-report/SKILL.md | 35 ---------- .../legal/planning-commission-brief/SKILL.md | 35 ---------- .../SKILL.md | 36 ---------- .../platus-kyc-document-collection/SKILL.md | 37 ---------- skills/legal/platus-notarization-api/SKILL.md | 36 ---------- .../SKILL.md | 31 --------- .../pretorin-nist-control-mapping/SKILL.md | 31 --------- .../SKILL.md | 34 --------- .../project-approval-risk-assessment/SKILL.md | 34 --------- .../legal/regulatory-controls-design/SKILL.md | 38 ---------- .../settleindex-casebot-risk-model/SKILL.md | 35 ---------- .../SKILL.md | 35 ---------- .../SKILL.md | 35 ---------- .../solomon-tax-preparation-workflow/SKILL.md | 69 ------------------- .../trustplane-compliance-audit/SKILL.md | 25 ------- skills/legal/trustplane-llm-safety/SKILL.md | 24 ------- .../trustplane-prompt-inspection/SKILL.md | 24 ------- .../legal/zoning-compliance-analysis/SKILL.md | 31 --------- 103 files changed, 3747 deletions(-) delete mode 100644 skills/legal/alphalaw-demand-letter-generation/SKILL.md delete mode 100644 skills/legal/alphalaw-medical-chronology/SKILL.md delete mode 100644 skills/legal/alphalaw-pi-case-workflow/SKILL.md delete mode 100644 skills/legal/anytime-agent-workflow/SKILL.md delete mode 100644 skills/legal/anytime-demand-letter/SKILL.md delete mode 100644 skills/legal/anytime-medical-chronology/SKILL.md delete mode 100644 skills/legal/automated-contract-summary/SKILL.md delete mode 100644 skills/legal/botmediation-adr-settlement/SKILL.md delete mode 100644 skills/legal/botmediation-negotiation-strategy/SKILL.md delete mode 100644 skills/legal/cardamon-gap-analysis/SKILL.md delete mode 100644 skills/legal/cardamon-horizon-scanning/SKILL.md delete mode 100644 skills/legal/cardamon-obligation-mapping/SKILL.md delete mode 100644 skills/legal/case-law-research/SKILL.md delete mode 100644 skills/legal/codex-billing-compliance-analysis/SKILL.md delete mode 100644 skills/legal/codex-claims-rebuttal-drafting/SKILL.md delete mode 100644 skills/legal/codex-medical-record-summarization/SKILL.md delete mode 100644 skills/legal/compli-gap-assessment/SKILL.md delete mode 100644 skills/legal/compliance-framework-mapping/SKILL.md delete mode 100644 skills/legal/construction-code-compliance/SKILL.md delete mode 100644 skills/legal/construction-contract-clause-analysis/SKILL.md delete mode 100644 skills/legal/construction-contract-review/SKILL.md delete mode 100644 skills/legal/construction-contract-risk-summary/SKILL.md delete mode 100644 skills/legal/construction-defect-analysis/SKILL.md delete mode 100644 skills/legal/construction-drawing-review/SKILL.md delete mode 100644 skills/legal/contract-leakage-analysis/SKILL.md delete mode 100644 skills/legal/contract-renewal-tracking/SKILL.md delete mode 100644 skills/legal/contract-review-triage/SKILL.md delete mode 100644 skills/legal/contract-term-extraction/SKILL.md delete mode 100644 skills/legal/covenant-fund-investment-review/SKILL.md delete mode 100644 skills/legal/covenant-mfns-election-analysis/SKILL.md delete mode 100644 skills/legal/covenant-nda-markup-triage/SKILL.md delete mode 100644 skills/legal/deepjudge-agile-workflows/SKILL.md delete mode 100644 skills/legal/deepjudge-knowledge-search/SKILL.md delete mode 100644 skills/legal/deepjudge-legal-intelligence-gathering/SKILL.md delete mode 100644 skills/legal/deposition-interrogation/SKILL.md delete mode 100644 skills/legal/deposition-searchable-record/SKILL.md delete mode 100644 skills/legal/deposition-video-analysis/SKILL.md delete mode 100644 skills/legal/dexterity-deal-intelligence-dashboard/SKILL.md delete mode 100644 skills/legal/dexterity-secure-contract-negotiation/SKILL.md delete mode 100644 skills/legal/dexterity-version-control-deals/SKILL.md delete mode 100644 skills/legal/dyspute-ai-mediation-assistant/SKILL.md delete mode 100644 skills/legal/dyspute-negotiation-proposal-generation/SKILL.md delete mode 100644 skills/legal/ecommerce-infringement-analysis/SKILL.md delete mode 100644 skills/legal/enforce-autonomous-takedown/SKILL.md delete mode 100644 skills/legal/expert-witness-matching/SKILL.md delete mode 100644 skills/legal/firstdrafts-case-summary/SKILL.md delete mode 100644 skills/legal/firstdrafts-judge-ruling-simulator/SKILL.md delete mode 100644 skills/legal/firstdrafts-litigation-motion/SKILL.md delete mode 100644 skills/legal/flipthrough-contract-proposal-analysis/SKILL.md delete mode 100644 skills/legal/flipthrough-negotiation-briefing/SKILL.md delete mode 100644 skills/legal/flipthrough-vendor-risk-assessment/SKILL.md delete mode 100644 skills/legal/genesis-case-evaluation-drafting/SKILL.md delete mode 100644 skills/legal/genesis-medical-chronology/SKILL.md delete mode 100644 skills/legal/genesis-tort-case-summary/SKILL.md delete mode 100644 skills/legal/habeas-case-strategy/SKILL.md delete mode 100644 skills/legal/habeas-citation-synthesis/SKILL.md delete mode 100644 skills/legal/habeas-custom-agent/SKILL.md delete mode 100644 skills/legal/haloo-brand-monitoring/SKILL.md delete mode 100644 skills/legal/haloo-portfolio-management/SKILL.md delete mode 100644 skills/legal/haloo-trademark-clearance/SKILL.md delete mode 100644 skills/legal/ip-portfolio-monitoring/SKILL.md delete mode 100644 skills/legal/lawxy-case-sentiment-analysis/SKILL.md delete mode 100644 skills/legal/lawxy-contract-review-studio/SKILL.md delete mode 100644 skills/legal/lawxy-due-diligence/SKILL.md delete mode 100644 skills/legal/lawxy-legal-agent-network/SKILL.md delete mode 100644 skills/legal/lawy-contract-intelligence/SKILL.md delete mode 100644 skills/legal/lawy-matter-summarization/SKILL.md delete mode 100644 skills/legal/lawy-verified-legal-research/SKILL.md delete mode 100644 skills/legal/legal-translation/SKILL.md delete mode 100644 skills/legal/legal1up-discovery/SKILL.md delete mode 100644 skills/legal/legal1up-joint-appendix/SKILL.md delete mode 100644 skills/legal/litigation-drafting/SKILL.md delete mode 100644 skills/legal/med-mal-case-evaluation/SKILL.md delete mode 100644 skills/legal/ml-contract-provision-extraction/SKILL.md delete mode 100644 skills/legal/pactum-negotiation-agent/SKILL.md delete mode 100644 skills/legal/pactum-price-list/SKILL.md delete mode 100644 skills/legal/patent-infringement-claim-chart/SKILL.md delete mode 100644 skills/legal/patent-infringement-detection/SKILL.md delete mode 100644 skills/legal/patent-invention-disclosure/SKILL.md delete mode 100644 skills/legal/patent-portfolio-analysis/SKILL.md delete mode 100644 skills/legal/patentability-assessment/SKILL.md delete mode 100644 skills/legal/patentprior-art-analysis/SKILL.md delete mode 100644 skills/legal/paxton-deposition-outline/SKILL.md delete mode 100644 skills/legal/paxton-legal-research-memo/SKILL.md delete mode 100644 skills/legal/paxton-policy-issue-memo/SKILL.md delete mode 100644 skills/legal/permitting-diligence-report/SKILL.md delete mode 100644 skills/legal/planning-commission-brief/SKILL.md delete mode 100644 skills/legal/platus-document-generation-agents/SKILL.md delete mode 100644 skills/legal/platus-kyc-document-collection/SKILL.md delete mode 100644 skills/legal/platus-notarization-api/SKILL.md delete mode 100644 skills/legal/pretorin-fedramp-compliance-drafting/SKILL.md delete mode 100644 skills/legal/pretorin-nist-control-mapping/SKILL.md delete mode 100644 skills/legal/pretorin-rmf-evidence-organization/SKILL.md delete mode 100644 skills/legal/project-approval-risk-assessment/SKILL.md delete mode 100644 skills/legal/regulatory-controls-design/SKILL.md delete mode 100644 skills/legal/settleindex-casebot-risk-model/SKILL.md delete mode 100644 skills/legal/settleindex-litigation-risk-modeling/SKILL.md delete mode 100644 skills/legal/settleindex-settlement-value-forecast/SKILL.md delete mode 100644 skills/legal/solomon-tax-preparation-workflow/SKILL.md delete mode 100644 skills/legal/trustplane-compliance-audit/SKILL.md delete mode 100644 skills/legal/trustplane-llm-safety/SKILL.md delete mode 100644 skills/legal/trustplane-prompt-inspection/SKILL.md delete mode 100644 skills/legal/zoning-compliance-analysis/SKILL.md diff --git a/skills/legal/alphalaw-demand-letter-generation/SKILL.md b/skills/legal/alphalaw-demand-letter-generation/SKILL.md deleted file mode 100644 index de4a3794..00000000 --- a/skills/legal/alphalaw-demand-letter-generation/SKILL.md +++ /dev/null @@ -1,33 +0,0 @@ ---- -name: alphalaw-demand-letter-generation -description: Generate comprehensive demand letters for personal injury claims using evidence synthesis and damages calculation -tags: [litigation, personal-injury, letter, drafting] -mode: drafting ---- - -# Demand Letter Generation - -Draft a compelling personal injury demand letter that presents liability, damages, and settlement expectations. - -## Prerequisites - -- Medical records showing injuries and treatment -- Incident documentation (police report, photos, witness statements) -- Economic damages documentation (medical bills, lost wages) -- Non-economic damages assessment (pain and suffering impact) -- Insurance policy information for at-fault party - -## Workflow - -1. **Evidence Synthesis** — Organize all medical records into a chronological timeline. Extract injury diagnoses, treatment dates, physician opinions on permanency. -2. **Liability Framework** — Structure the narrative of how the incident occurred, establish duty and breach, and link to injuries causally. -3. **Damages Calculation** — Sum all economic damages (medical expenses, lost income, future care costs). Calculate non-economic damages using appropriate multipliers or per-diem method. -4. **Demand Letter Drafting** — Write a professional letter: incident description, liability basis, comprehensive damages schedule, specific demand amount, and settlement deadline. - -## Pitfalls - -- Over-demanding damages reduces settlement credibility — anchor to provable damages -- Missing medical connection between incident and specific injuries weakens the claim -- Don't reveal settlement floor or minimum acceptance amount -- Include a clear deadline (typically 30 days) to create negotiation urgency -- Avoid emotional language — stick to facts and documented damages diff --git a/skills/legal/alphalaw-medical-chronology/SKILL.md b/skills/legal/alphalaw-medical-chronology/SKILL.md deleted file mode 100644 index e1421340..00000000 --- a/skills/legal/alphalaw-medical-chronology/SKILL.md +++ /dev/null @@ -1,32 +0,0 @@ ---- -name: alphalaw-medical-chronology -description: Build structured medical chronologies from treatment records for PI case evaluation and expert presentation -tags: [litigation, personal-injury, summary, analysis] -mode: analysis ---- - -# Medical Chronology Builder - -Organize scattered medical records into a clear chronological timeline showing injury progression, treatment, and causation. - -## Prerequisites - -- All available medical records (ER visits, hospital admissions, specialist consultations, physical therapy notes) -- Incident documentation establishing the triggering event -- Pre-incident medical history (if available) to establish baseline health - -## Workflow - -1. **Record Collection** — Gather all medical documentation. Identify gaps in the treatment timeline. Request missing records from providers. -2. **Data Extraction** — For each medical encounter extract: date, provider/facility, diagnosis, treatment rendered, medications prescribed, follow-up instructions. -3. **Chronology Construction** — Arrange entries chronologically. Highlight causation links between the incident and each diagnosed injury. -4. **Gap Analysis** — Identify treatment gaps, unexpected delays, or inconsistencies that may weaken the case. Flag for attorney review. -5. **Summary Report** — Produce a one-page executive summary of injury severity, treatment completeness, and permanency assessments. - -## Pitfalls - -- Missing follow-up visits create treatment gaps opponents will exploit -- Provider opinions on causation must be explicit — don't assume -- Pre-existing conditions must be documented and distinguished from new injuries -- Ensure all entries have dates — undated records are largely useless -- Include medication lists to show treatment intensity and continuity diff --git a/skills/legal/alphalaw-pi-case-workflow/SKILL.md b/skills/legal/alphalaw-pi-case-workflow/SKILL.md deleted file mode 100644 index 113623d1..00000000 --- a/skills/legal/alphalaw-pi-case-workflow/SKILL.md +++ /dev/null @@ -1,31 +0,0 @@ ---- -name: alphalaw-pi-case-workflow -description: AI-powered personal injury case management workflow — intake triage, case evaluation, pre-litigation tracking, and litigation scheduling -tags: [litigation, personal-injury, intake, checklist] -mode: drafting ---- - -# PI Case Management Workflow - -Manage a personal injury case from initial intake through litigation using AI-assisted triage and structured workflows. - -## Prerequisites - -- Client intake form with incident details, injuries, treatment history -- Insurance information (at-fault party, policy limits) -- Incident date, jurisdiction, applicable statutes of limitation - -## Workflow - -1. **Intake Triage** — Extract key facts from intake form: incident type, injuries claimed, at-fault party, insurance details. Qualify the case using viability checklist (liability clarity, damages magnitude, jurisdiction favorable rules). -2. **Pre-Litigation Phase** — Build case timeline, identify all responsible parties, track medical treatment milestones, schedule deadlines. -3. **Demand Phase** — Draft demand package using evidence gathered (medical records, liability documents, damages calculation). -4. **Litigation Phase** — Create litigation calendar with all court deadlines, discovery deadlines, and settlement negotiation checkpoints. - -## Pitfalls - -- Don't assume liability — verify all elements of the specific cause of action -- Missed statute of limitations is case-ending — always calculate first -- Insurance policy limits may cap recoverable amounts — check early -- Multiple at-fault parties may shift liability — identify all defendants before demand -- Jurisdiction-specific rules for comparative negligence affect strategy diff --git a/skills/legal/anytime-agent-workflow/SKILL.md b/skills/legal/anytime-agent-workflow/SKILL.md deleted file mode 100644 index 0256dee3..00000000 --- a/skills/legal/anytime-agent-workflow/SKILL.md +++ /dev/null @@ -1,47 +0,0 @@ ---- -name: plaintiff-case-agentic-workflow -language: en -description: >- - Orchestrates an end-to-end agentic workflow for plaintiff complex - litigation: intake triage, investigation, discovery review, and trial - prep. Use when managing a full plaintiff case lifecycle with AI agents - that analyze records, surface case theory, and automate repetitive tasks. - Trigger keywords: plaintiff workflow, case triage, agentic litigation, - case automation, nursing home case, medmal workflow, verdict preparation. -tags: - - litigation - - analysis - - personal-injury ---- - -# Plaintiff Case Agentic Workflow - -Coordinates multi-agent case management across intake through verdict. - -## Prerequisites -- Intake form or initial client data -- Uploaded case documents (records, complaints, correspondence) -- Practice area classification (nursing home, medmal, PI, etc.) - -## Workflow -1. **Intake & Triage Agent** — Assess case strength, identify facts, - create encrypted client profile, flag conflicts, assign priority. -2. **Investigation Agent** — Analyze uploaded records, build initial - medical chronology, identify liable parties and potential defendants. -3. **Discovery Agent** — Review voluminous discovery, surface - inconsistencies, flag critical evidence, build deposition outlines. -4. **Strategy Agent** — Synthesize findings into case theory, - recommend damages range, identify expert needs, prepare trial timeline. -5. **Resolution Agent** — Draft demand letters, negotiate settlement - terms, or prepare courtroom exhibits based on case stage. -6. **Quality Check** — Attorney review gate before any external - communication. Verify all facts, citations, and calculations. - -## Pitfalls -- Maintain data encryption at every agent handoff — never expose - client data across agent boundaries. -- Zero data training: ensure no case data leaves the firm's encrypted - environment for model training. -- Flag jurisdiction-specific rules (statutes of limitations, notice - requirements, damage caps) at intake. -- Always require attorney sign-off before demand letters or filings. diff --git a/skills/legal/anytime-demand-letter/SKILL.md b/skills/legal/anytime-demand-letter/SKILL.md deleted file mode 100644 index 92fd3f3b..00000000 --- a/skills/legal/anytime-demand-letter/SKILL.md +++ /dev/null @@ -1,46 +0,0 @@ ---- -name: demand-letter-generator -language: en -description: >- - Drafts persuasive, case-specific demand letters grounded in facts, - medical chronology, and damages calculation. Use when preparing pre-suit - demand packages for PI, medmal, nursing home, or personal injury cases. - Trigger keywords: demand letter, demand package, settlement demand, pre-suit - demand, demand drafting, claim demand, settlement letter. -tags: - - drafting - - litigation - - personal-injury ---- - -# Demand Letter Generator - -Produces a persuasive, fact-anchored demand letter for plaintiff cases. - -## Prerequisites -- Medical chronology or clinical overview -- Documented damages (medical bills, lost wages, pain/suffering) -- Liability facts and responsible parties -- Jurisdiction (for statutory caps, notice requirements) - -## Workflow -1. **Frame the Narrative** — Open with the incident, establish liability, - and introduce the injured party with empathy. -2. **Present the Facts** — Summarize the chronology of events, - treatments, and injuries without excessive detail. -3. **Quantify Damages** — Itemize medical expenses, lost wages, - property damage, and present a calculated pain/suffering figure. -4. **Establish Causation** — Link injuries directly to the incident - using medical evidence and expert quotes where available. -5. **State the Demand** — Clearly state the settlement amount and - deadline for response. Reference jurisdiction-specific notice - requirements if applicable. -6. **Attach Evidence** — Reference enclosures: medical records summary, - bill summaries, wage documentation, photos. - -## Pitfalls -- Never admit comparative fault unless strategy requires it. -- Avoid overly emotional language — persuasive but professional tone. -- Include response deadline (typically 30 days). -- Check for jurisdiction-specific demand letter formatting rules. -- Do not reveal internal calculations — present the bottom line. diff --git a/skills/legal/anytime-medical-chronology/SKILL.md b/skills/legal/anytime-medical-chronology/SKILL.md deleted file mode 100644 index 5b4d9ab0..00000000 --- a/skills/legal/anytime-medical-chronology/SKILL.md +++ /dev/null @@ -1,49 +0,0 @@ ---- -name: medical-chronology-builder -language: en -description: >- - Transforms voluminous medical records into structured, cited timelines - mapping causation, standard-of-care deviations, and damages. Use when - plaintiff firms need medical chronologies from EHR, facility notes, - treatment records, or pharmacy data for PI, medmal, nursing home, or - personal injury cases. Trigger keywords: medical chronology, treatment - timeline, causation timeline, medical overview, record review, standard - of care deviation, clinical events timeline. -tags: - - litigation - - summarization - - personal-injury ---- - -# Medical Chronology Builder - -Produces a structured, citation-linked medical chronology from raw clinical -records suitable for demand letters, discovery, and trial preparation. - -## Prerequisites -- Uploaded medical records (EHR, facility notes, pharmacy logs, imaging reports) -- Case type context (PI, medmal, nursing home, workers comp, etc.) -- Key dates: incident date, diagnosis dates, treatment start/end - -## Workflow -1. **Ingest & Classify** — Sort records by type (EHR, nursing notes, - pharmacy, imaging, discharge summaries) and date. -2. **Identify Key Events** — Extract diagnoses, treatments, provider - actions, test results, and clinical deviations from standard of care. -3. **Map Causation Chain** — Link each event to the incident, show - gaps in treatment, missed care, or deviations. -4. **Build Cited Timeline** — Create a date-ordered chronology with - inline citations to source documents (page numbers, document IDs). -5. **Flag Litigation Insights** — Highlight standard-of-care breaches, - medication errors, falls, pressure ulcers, dehydration, delayed - treatment, and systemic neglect patterns. -6. **Produce Output** — Generate a clean, attorney-ready chronology - suitable for expert review and admission into demand packages. - -## Pitfalls -- Do NOT infer causation beyond what records support — mark gaps as - "insufficient records" not "no negligence." -- Nursing home cases require ownership/structure timeline alongside - clinical chronology. -- Always preserve original document citations for admissibility. -- Flag conflicting dates between providers for attorney review. diff --git a/skills/legal/automated-contract-summary/SKILL.md b/skills/legal/automated-contract-summary/SKILL.md deleted file mode 100644 index 244743ac..00000000 --- a/skills/legal/automated-contract-summary/SKILL.md +++ /dev/null @@ -1,38 +0,0 @@ -name: automated-contract-summary -language: en -description: Generates structured executive summaries of contracts using ML — captures key terms, party obligations, risk allocations, and compliance requirements in a standardized format. Optimized for high-volume review where speed and consistency matter. -tags: - - summarization - - agreement - - corporate ---- - -# Automated Contract Summarization - -Produces standardized executive summaries of contracts using machine learning, capturing essential terms and obligations for rapid review and comparison across portfolios. - -## Prerequisites - -Before executing, collect: - -1. **Contract document** — full agreement text (PDF, DOCX, or extracted text) -2. **Summary template** — standard fields expected (parties, effective date, term, key obligations, termination, liability, etc.) -3. **Audience** — executive summary (high-level), legal review (detailed), or operational (action items) -4. **Volume** — single document or batch processing - -## Workflow - -1. **Document parsing** — extract clean text from contract, handle tables and multi-column layouts -2. **Section identification** — map document to standard contract sections (parties, term, obligations, etc.) -3. **Key term extraction** — pull out dates, amounts, thresholds, named entities using ML models -4. **Obligation identification** — extract affirmative and negative obligations for each party -5. **Risk flagging** — highlight unusual, asymmetric, or high-risk provisions (uncapped liability, auto-renewal) -6. **Summary assembly** — compose structured summary following the target template -7. **Validation** — cross-check extracted terms against original document, flag uncertain extractions - -## Pitfalls - -- Summaries are not substitutes for full review — legal conclusions require human judgment -- Ambiguous clauses may be summarized incorrectly — flag for manual review -- Multi-party contracts increase complexity — obligations must be attributed to correct parties -- Boilerplate extraction (governing law, notice provisions) should be templated, not free-form diff --git a/skills/legal/botmediation-adr-settlement/SKILL.md b/skills/legal/botmediation-adr-settlement/SKILL.md deleted file mode 100644 index c61b3db2..00000000 --- a/skills/legal/botmediation-adr-settlement/SKILL.md +++ /dev/null @@ -1,33 +0,0 @@ ---- -name: botmediation-adr-settlement -description: Analyze alternative dispute resolution outcomes and prepare settlement proposals for ADR proceedings -tags: [litigation, mediation, summary, analysis] -mode: analysis ---- - -# ADR Settlement Analysis - -Analyze dispute resolution outcomes and prepare structured settlement proposals for alternative dispute resolution proceedings. - -## Prerequisites - -- Dispute description and relevant facts -- Both parties' positions and demands -- Applicable law and relevant precedent -- Prior settlement discussions or mediation attempts - -## Workflow - -1. **Dispute Mapping** — Document the core issues in dispute, identify negotiable vs. non-negotiable positions, and map the dispute landscape. -2. **Valuation Analysis** — Calculate fair settlement value considering risk of litigation, cost savings, timeline to resolution, and relationship preservation. -3. **Settlement Proposal Drafting** — Create a structured proposal with specific terms, payment schedule (if applicable), and release language. -4. **Outcome Comparison** — Compare settlement outcomes against potential trial/arbitration results including risk factors, costs, and timeline. -5. **Post-Settlement Documentation** — Draft settlement agreement, release of claims, and any confidentiality or non-disparagement provisions. - -## Pitfalls - -- ADR proposals must be specific — vague terms lead to unresolved disputes -- Include enforcement mechanisms in settlement agreements -- Confidentiality terms may limit future precedent — advise client on tradeoffs -- Tax implications of structured settlements should be reviewed -- Ensure all parties with legal standing are included in releases diff --git a/skills/legal/botmediation-negotiation-strategy/SKILL.md b/skills/legal/botmediation-negotiation-strategy/SKILL.md deleted file mode 100644 index 71a79da3..00000000 --- a/skills/legal/botmediation-negotiation-strategy/SKILL.md +++ /dev/null @@ -1,33 +0,0 @@ ---- -name: botmediation-negotiation-strategy -description: Prepare AI-guided mediation negotiation strategy with settlement positioning and concession planning -tags: [litigation, mediation, analysis, drafting] -mode: analysis ---- - -# Mediation Negotiation Strategy - -Develop a structured negotiation strategy for mediation proceedings, including position setting, concession planning, and settlement readiness. - -## Prerequisites - -- Case strength assessment (liability analysis, damages estimate) -- Opposing party's demand or offer (if available) -- Client's optimal outcome, target range, and walk-away point -- Relevant case law and jurisdictional precedent for similar disputes - -## Workflow - -1. **Case Valuation** — Establish a defensible case value range using damages documentation and comparable case outcomes. Calculate best, target, and worst cases. -2. **Position Strategy** — Set the opening demand/offer based on case strength, client objectives, and negotiation leverage. -3. **Concession Planning** — Map out concession moves: size, timing, and conditional language for each step toward settlement. -4. **Mediation Brief** — Draft a mediation brief summarizing liability, damages, key evidence, and valuation rationale for the mediator. -5. **Negotiation Playbook** — Create a running checklist tracking offers, counteroffers, and mediator feedback during the session. - -## Pitfalls - -- Opening positions that are too extreme can shut down productive negotiation -- Don't reveal your walk-away point or bottom line prematurely -- Mediators respond to well-documented valuations — bring evidence, not just demands -- Cultural and relationship dynamics affect negotiation — adapt tone accordingly -- Document all settlement discussions for fee approval and client records diff --git a/skills/legal/cardamon-gap-analysis/SKILL.md b/skills/legal/cardamon-gap-analysis/SKILL.md deleted file mode 100644 index 2c9ec341..00000000 --- a/skills/legal/cardamon-gap-analysis/SKILL.md +++ /dev/null @@ -1,56 +0,0 @@ ---- -name: gap-analysis -language: en -description: >- - Compares an organization's existing compliance controls, policies, and - procedures against extracted regulatory obligations to identify coverage - gaps. Produces a remediation plan with prioritized actions. Use when - assessing compliance maturity or preparing for regulatory audits. -tags: - - regulatory - - compliance - - analysis ---- - -# Gap Analysis - -Systematic comparison of existing compliance posture against regulatory -obligations to identify, prioritize, and remediate coverage gaps. - -## Prerequisites - -- Extracted obligation register (from obligation-mapping) -- Inventory of existing policies, procedures, controls -- Risk methodology framework (existing or standard industry framework) -- Prior gap analysis results (if iterative) - -## Workflow - -1. **Control inventory** — catalog all existing compliance artifacts: - - Policies and procedures (with effective dates and owners) - - Technical controls (system configurations, access controls) - - Training programs and completion rates - - Monitoring and reporting mechanisms -2. **Mapping** — align each control against its corresponding obligation(s): - - Direct match: control explicitly addresses the obligation - - Partial match: control addresses part but not all elements - - Gap: no corresponding control identified -3. **Scoring** — for each mapping: - - Coverage completeness (full/partial/none) - - Recency (control updated within last 12 months?) - - Evidence quality (documented, tested, audited) -4. **Remediation planning** — for each gap: - - Classify remediation type (new control, modify existing, augment) - - Estimate effort (person-weeks) - - Assign priority based on obligation severity and closure timeline - - Draft the remediation action with acceptance criteria -5. **Reporting** — produce gap analysis report with executive summary, - detailed findings by obligation, and remediation roadmap - -## Pitfalls - -- A control that exists on paper but is not tested or enforced is functionally a gap — verify operational effectiveness -- Don't ignore implicit controls: informal practices and tribal knowledge matter, document them -- Gap analyses are snapshots — establish a periodic cadence, not one-off exercises -- Be careful with third-party controls: vendors may provide compliance evidence but it doesn't transfer liability -- Over-mapping is the inverse risk: claiming a control covers an obligation when it actually covers a different one diff --git a/skills/legal/cardamon-horizon-scanning/SKILL.md b/skills/legal/cardamon-horizon-scanning/SKILL.md deleted file mode 100644 index 7644ec7f..00000000 --- a/skills/legal/cardamon-horizon-scanning/SKILL.md +++ /dev/null @@ -1,53 +0,0 @@ ---- -name: horizon-scanning -language: en -description: >- - Continuously monitors regulatory landscapes for changes relevant to a - specific business. Ingests global regulatory updates, filters by relevance, - summarizes impact, and produces an actionable change advisory. Use when - tracking regulatory developments affecting a particular product or market. -tags: - - regulatory - - compliance - - research ---- - -# Horizon Scanning - -Structured process for detecting, filtering, and analyzing regulatory changes -that may affect a specific business. - -## Prerequisites - -- Business profile (products, services, jurisdictions, activity types) -- Source regulatory feeds (government gazettes, agency calendars, monitors) -- Time window (rolling 30 days, quarterly, ad-hoc) - -## Workflow - -1. **Signal collection** — aggregate regulatory signals from defined sources: - - New legislation, regulations, directives - - Agency guidance, interpretive letters, FAQs - - Enforcement actions and consent orders - - Draft regulations and proposed rules -2. **Filtering** — apply relevance filters: - - Jurisdiction overlap with business operations - - Sector/domain match (products/services engaged) - - Effective date window (active or upcoming within 12 months) -3. **Impact assessment** — for each relevant signal: - - Summarize the change in plain language - - Identify affected obligations (new, modified, repealed) - - Assess materiality (high/medium/low) based on operational impact -4. **Advisory output** — produce a structured update with: - - Regulation identifier, effective date, source - - Business-specific impact statement - - Recommended action items and responsible parties - - Links to full text and comment periods - -## Pitfalls - -- Don't over-filter: early-stage draft regulations can still signal enforcement priorities -- Enforcement actions, while not new rules, often reveal how agencies interpret existing obligations -- International regulations in translation may lose nuance — flag and note the original language -- Multiple overlapping regulations may address the same issue — identify the controlling hierarchy -- Comment periods for draft regulations are compliance-critical opportunities — never miss deadlines diff --git a/skills/legal/cardamon-obligation-mapping/SKILL.md b/skills/legal/cardamon-obligation-mapping/SKILL.md deleted file mode 100644 index 41ed9255..00000000 --- a/skills/legal/cardamon-obligation-mapping/SKILL.md +++ /dev/null @@ -1,46 +0,0 @@ ---- -name: obligation-mapping -language: en -description: >- - Extracts regulatory obligations from dense regulations across jurisdictions. - Breaks down multi-level regulations into clear article-level obligations, - classifies applicability to a business, and prioritizes by risk level. - Use when translating regulations into actionable compliance requirements. -tags: - - regulatory - - compliance - - analysis ---- - -# Obligation Mapping - -Systematic process for extracting, classifying, and prioritizing regulatory obligations from complex regulatory text. - -## Prerequisites - -- Source regulation text (statute, directive, regulation, guidance) -- Target business description (products, activities, markets, jurisdictions) -- Jurisdiction and effective dates - -## Workflow - -1. **Regulation ingestion** — identify the full regulatory text including all levels (law, regulation, directive, guidance) -2. **Obligation extraction** — parse each article/section for mandatory language (shall, must) vs. permissive language (may) vs. aspirational (should) -3. **Applicability scoring** — match extracted obligations against the target business profile: - - Product/Service fit: does the business offer regulated products or services? - - Activity match: does the business engage in covered activities? - - Geographic scope: are operations in covered jurisdictions? - - Entity type: does the entity type fall under the regulation? -4. **Risk prioritization** — assign priority (critical/high/medium/low) based on: - - Enforcement severity (fines, criminal, license revocation) - - Probability of enforcement action - - Remediation complexity and cost -5. **Output format** — produce a structured obligation register with article reference, obligation text, applicability rationale, priority, and responsible function - -## Pitfalls - -- Don't conflate permissive guidance with mandatory obligation — always flag the exact statutory language -- Watch for nested obligations: an article may reference another regulation for the actual requirement -- International regulations often have implementation deadlines that differ from entry-into-force dates -- Cross-references between regulations can create cascading obligations — trace the full chain -- Translated regulations may use ambiguous terminology — note language version and flag uncertain translations diff --git a/skills/legal/case-law-research/SKILL.md b/skills/legal/case-law-research/SKILL.md deleted file mode 100644 index c54188d3..00000000 --- a/skills/legal/case-law-research/SKILL.md +++ /dev/null @@ -1,29 +0,0 @@ ---- -name: Case Law Research -description: Research and analyze binding case law precedents with proper citations, jurisdictional specificity, and analogous fact patterns. Identifies controlling authority and distinguishing cases. -tags: - - practice_area: litigation - - document_type: memo, brief, analysis - - skill_mode: research ---- - -# Case Law Research - -## Prerequisites -- Legal issue or question of law -- Jurisdiction (court level, geographic) -- Timeframe for precedents (optional) - -## Workflow -1. **Identify the legal question** — narrow the issue to a specific, researchable proposition -2. **Search case law corpus** — find binding authority and persuasive precedents -3. **Verify citations** — confirm citations are current, not overruled, and properly formatted -4. **Analyze holdings** — extract key holdings, reasoning, and factual contexts -5. **Map analogies** — identify cases with analogous facts and compare outcomes -6. **Flag weaknesses** — note distinguishing factors, contrary authority, or circuit splits - -## Pitfalls -- Always verify citation validity; do not rely on unverified references -- Distinguish between binding and persuasive authority -- Check for subsequent treatment (overruled, distinguished, followed) -- Circuit splits require noting divergent approaches diff --git a/skills/legal/codex-billing-compliance-analysis/SKILL.md b/skills/legal/codex-billing-compliance-analysis/SKILL.md deleted file mode 100644 index 56d6e26e..00000000 --- a/skills/legal/codex-billing-compliance-analysis/SKILL.md +++ /dev/null @@ -1,37 +0,0 @@ -# Healthcare Billing Compliance Risk Assessment & Audit Defense - -## Prerequisites -- Billing records and claim data (superbills, CMS-1500, UB-04) -- Payer contract terms and fee schedules -- Audit notice or scope letter (if under audit) - -## Workflow -1. Analyze billing data against applicable standards: - - CMS benefit parameters (medically necessary, reasonable and customary) - - Payer-specific coverage policies and LCDs/NCDs - - State Medicaid regulations and managed care rules -2. Cross-reference CPT codes with medical records for each claim: - - Documentation matches billed service? - - Modifier accuracy and medical necessity support? - - Day/period-of-service alignment? -3. Calculate risk exposure: - - Likely overpayments by claim type and service - - Penalty multipliers (False Claims Act treble damages, civil monetary penalties) - - Provider-level patterns (outliers, consistent practices vs. errors) -4. Generate audit defense positions: - - Documented medical necessity arguments - - Good-faith coding practices (training records, CMS manual citations) - - Distinguish errors from willful misrepresentation -5. Produce a risk assessment report ranked by exposure severity - -## Pitfalls -- False Claims Act requires "knowing" violations — document good-faith coding practices -- Lookback periods vary: voluntary disclosure = 3 years, OIG = 6 years, FCA = 6 years -- LCDs are jurisdiction-specific — verify the correct MAC region -- Self-disclosure protocol (SDP) timing affects penalty severity -- CMS recoupment rules differ for Medicare Advantage vs. Traditional Medicare - -## Tags -practice_area: regulatory -document_type: analysis, checklist -skill_mode: analysis diff --git a/skills/legal/codex-claims-rebuttal-drafting/SKILL.md b/skills/legal/codex-claims-rebuttal-drafting/SKILL.md deleted file mode 100644 index 510c1b5c..00000000 --- a/skills/legal/codex-claims-rebuttal-drafting/SKILL.md +++ /dev/null @@ -1,37 +0,0 @@ -# Medical Billing Rebuttal Letter & CMS Guidance Citation Drafting - -## Prerequisites -- Carrier denial letter or recoupment notice -- Specific claim(s) in dispute with CPT/ICD-10 codes -- Relevant CMS manual sections, LCDs, NCDs, or carrier guidelines -- Patient chart documentation supporting medical necessity - -## Workflow -1. Parse the denial/recoupment notice to identify: - - Specific codes/services disputed - - Carrier's stated rationale (medical necessity, coding error, documentation gap) - - Appeal deadline and required format -2. For each disputed claim: - - Pull the applicable CMS manual section, LCD, or NCD - - Extract the patient's clinical documentation that supports medical necessity - - Identify discrepancies between the carrier's analysis and the clinical record -3. Draft the rebuttal letter containing: - - Clear identification of each disputed service - - Citation to the specific guideline paragraph or section number - - Clinical evidence from records that satisfies the guideline requirement - - Explanation of any modifiers used and why they are correct - - Request for specific action (reopen, pay, reconsider) -4. Create a coverage checklist showing each disputed item and its supporting citation -5. Output a polished rebuttal letter ready for attorney signature - -## Pitfalls -- Deadlines are strict — track appeal windows by payer (30–180 days typical) -- CMS manual section numbers changed in recent years — cite the current version -- LCDs are binding on MAC decisions — NCDs are national but not always followed -- Don't cite carrier policies as if they're CMS regulations — maintain hierarchy -- Rebuttal must address each denial reason individually — group denials need separate analysis - -## Tags -practice_area: regulatory -document_type: letter -skill_mode: drafting diff --git a/skills/legal/codex-medical-record-summarization/SKILL.md b/skills/legal/codex-medical-record-summarization/SKILL.md deleted file mode 100644 index 82f88690..00000000 --- a/skills/legal/codex-medical-record-summarization/SKILL.md +++ /dev/null @@ -1,32 +0,0 @@ -# Medical Record Summarization for Billing Defense - -## Prerequisites -- Patient medical records (encounter notes, orders, labs, imaging reports) -- Applicable CPT codes and billing period -- Carrier/payer guidelines (CMS manuals, state Medicaid, commercial plan rules) - -## Workflow -1. Ingest medical records and index by date, provider, service type -2. For each encounter: - - Extract diagnoses (ICD-10), procedures (CPT/HCPCS), modifiers - - Map services to clinical guidelines and coverage determinations - - Identify documentation gaps (missing medical necessity, lack of supporting notes) - - Flag encounters with coding inconsistencies (upcoding, unbundling signals) -3. Generate an encounter-by-encounter summary: - - Clinical narrative in plain language - - Coding compliance assessment per payer guidelines - - Supporting evidence citations (CMS manual sections, carrier bulletins) -4. Produce an overall audit summary: risk score, recovery potential, top issue categories -5. Output attorney-ready summaries with cite-ready references - -## Pitfalls -- CMS guidelines change annually — verify the applicable year's manuals -- Modifier usage (25, 59, X{EPSU}) must match current AMA conventions -- Medical necessity determinations require specific documentation — vague notes = denial risk -- Don't conflate professional vs. facility billing rules — they differ by setting -- Coding guidelines (ICD-10-CM, CPT, HCPCS) have different versioning cycles - -## Tags -practice_area: regulatory -document_type: summary, analysis -skill_mode: summarization diff --git a/skills/legal/compli-gap-assessment/SKILL.md b/skills/legal/compli-gap-assessment/SKILL.md deleted file mode 100644 index beb4cd6f..00000000 --- a/skills/legal/compli-gap-assessment/SKILL.md +++ /dev/null @@ -1,38 +0,0 @@ -name: compliance-gap-assessment -language: en -description: Conducts AI-powered gap assessments between an organization's current compliance posture and target regulatory frameworks — GDPR, SOC 2, NIST 800-53, ISO 27001. Produces risk-scored findings with remediation priorities and evidence requirements for each gap. -tags: - - analysis - - checklist - - regulatory ---- - -# AI-Powered Compliance Gap Assessment - -Automates the identification and scoring of compliance gaps between an organization's current state and target regulatory or standards frameworks, producing prioritized remediation roadmaps. - -## Prerequisites - -Before executing, collect: - -1. **Target framework(s)** — GDPR, SOC 2, NIST 800-53, ISO 27001, HIPAA, PCI-DSS, etc. -2. **Organization profile** — industry, jurisdiction, company size, data processing activities -3. **Current documentation** — existing policies, procedures, controls inventory, audit reports -4. **Scope boundaries** — systems, departments, or processes in scope - -## Workflow - -1. **Framework ingestion** — load target regulatory requirements or standards clauses -2. **Profile mapping** — map organization's activities to applicable regulatory sections -3. **Evidence review** — analyze existing policies, procedures, and control documentation -4. **Gap identification** — flag missing controls, inadequate policies, or non-conforming practices -5. **Risk scoring** — assign risk level to each gap based on regulatory severity and likelihood -6. **Remediation planning** — suggest specific controls, policy changes, or process improvements -7. **Report generation** — produce structured gap assessment with findings, scores, and action items - -## Pitfalls - -- Frameworks overlap significantly — don't treat gaps in isolation when controls cross-framework -- Self-assessments tend to overstate compliance — require independent validation -- Regulatory changes are continuous — gaps identified today may shift with new guidance -- Some requirements have jurisdictional nuances not captured by generic frameworks diff --git a/skills/legal/compliance-framework-mapping/SKILL.md b/skills/legal/compliance-framework-mapping/SKILL.md deleted file mode 100644 index 167a328b..00000000 --- a/skills/legal/compliance-framework-mapping/SKILL.md +++ /dev/null @@ -1,38 +0,0 @@ -name: compliance-framework-mapping -language: en -description: Maps regulatory requirements across multiple frameworks (GDPR, SOC 2, NIST, ISO 27001, etc.) to identify overlaps, gaps, and dual-fulfillment opportunities. Produces a cross-reference matrix enabling efficient compliance with simultaneous standards. -tags: - - research - - memo - - regulatory ---- - -# Cross-Framework Regulatory Mapping - -Identifies overlaps, gaps, and dual-fulfillment opportunities across multiple regulatory frameworks to streamline compliance programs and reduce redundant controls. - -## Prerequisites - -Before executing, collect: - -1. **Applicable frameworks** — list of all regulations/standards the organization must comply with -2. **Jurisdictions** — countries/regions where the organization operates or processes data -3. **Industry sector** — healthcare, finance, SaaS, retail, etc. (affects applicable regulations) -4. **Data classification** — types of data handled (PII, financial, health, IP, etc.) - -## Workflow - -1. **Framework ingestion** — load requirements from each applicable framework/standard -2. **Requirement normalization** — express each requirement in a common reference format -3. **Overlap detection** — identify requirements across frameworks that address the same obligation -4. **Gap identification** — flag regulatory obligations not covered by existing controls -5. **Dual-fulfillment matrix** — produce cross-reference showing which controls satisfy which frameworks -6. **Priority ranking** — score by regulatory severity, audit frequency, and enforcement risk -7. **Roadmap output** — generate phased implementation plan addressing highest-impact requirements first - -## Pitfalls - -- Similar wording ≠ same requirement — check scope and exceptions carefully -- Framework updates are asynchronous — a mapping today may break with next revision -- Some frameworks have jurisdictional clauses that don't apply universally -- Over-optimizing for dual-fulfillment can create blind spots in framework-specific requirements diff --git a/skills/legal/construction-code-compliance/SKILL.md b/skills/legal/construction-code-compliance/SKILL.md deleted file mode 100644 index 24aa6ae9..00000000 --- a/skills/legal/construction-code-compliance/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ ---- -name: construction-code-compliance -description: Verify construction plans against building codes, fire safety, ADA, and accessibility requirements -tags: - - practice: regulatory - - document: checklist - - mode: analysis ---- - -# Construction Code Compliance Checklist - -Generate comprehensive code compliance checklists for construction projects. - -## Prerequisites -- Project type and occupant classification -- Building height, area, and construction type -- Applicable code cycle (e.g., IBC 2021, NFPA 101, ADA 2010) -- Local amendments or jurisdiction-specific requirements - -## Workflow -1. **Identify**: Determine building classification and applicable code sections -2. **Generate Checklist**: Create room-by-room and system-by-system compliance items: - - Egress (capacity, distance, width, signage) - - Fire protection (sprinklers, alarms, ratings) - - Accessibility (routes, fixtures, signage) - - Energy and insulation - - Structural provisions -3. **Flag Gaps**: Highlight areas where drawings don't clearly address code requirements -4. **Recommend**: Suggest plan revisions to achieve compliance - -## Pitfalls -- Local amendments override model codes — always check first -- Occupant classification drives most requirements — get it right first -- Remember life safety is separate from accessibility requirements -- Energy codes vary widely by climate zone — verify the correct one diff --git a/skills/legal/construction-contract-clause-analysis/SKILL.md b/skills/legal/construction-contract-clause-analysis/SKILL.md deleted file mode 100644 index 5f27ed70..00000000 --- a/skills/legal/construction-contract-clause-analysis/SKILL.md +++ /dev/null @@ -1,36 +0,0 @@ ---- -name: construction-contract-clause-analysis -description: Deep-dive analysis of specific construction contract clauses with playbook-based guidance -tags: - practice: transactional - document: analysis - mode: analysis ---- - -# Construction Contract Clause Analysis - -Deep analysis of individual construction contract clauses using industry playbook standards. - -## Prerequisites - -- Specific clause text to analyze -- Applicable industry standard (consensus docs, AIA, EJCDC, custom) -- Client's negotiation position (aggressive, moderate, defensive) -- Relevant project context - -## Workflow - -1. **Clause classification** — Identify clause type (indemnity, insurance, change order, schedule, payment, termination, dispute resolution) -2. **Standard comparison** — Compare against industry standard language for the clause type -3. **Risk assessment** — Evaluate risk allocation using construction-specific criteria -4. **Playbook matching** — Find client's preferred/fallback language in playbook library -5. **Negotiation recommendations** — Suggest revision language aligned with client's position -6. **Video reference** — Link to relevant clause guidance content (clause-linked training) - -## Pitfalls - -- Construction indemnity clauses are jurisdiction-specific — verify enforceability in the project state -- Flow-down clauses create cascading obligations — trace through the entire chain -- Change order pricing terms (GMP vs. cost-plus vs. lump sum) fundamentally change risk -- Pay-when-paid vs. pay-if-paid distinction is critical and varies by state -- Stop-work and suspension clauses often hide schedule damage exposure diff --git a/skills/legal/construction-contract-review/SKILL.md b/skills/legal/construction-contract-review/SKILL.md deleted file mode 100644 index 16e27e2a..00000000 --- a/skills/legal/construction-contract-review/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ ---- -name: construction-contract-review -description: Review construction contracts (prime + subcontract) for risk, liability, and deviation from standards -tags: - practice: transactional - document: agreement - mode: analysis ---- - -# Construction Contract Review - -Systematic analysis of prime contracts and subcontracts for construction industry risk allocation. - -## Prerequisites - -- Full contract text (prime contract or subcontract) -- Client's contract standards/playbook (preferred + fallback language by clause type) -- Project role context (GC, subcontractor, owner, AE, owner's rep) - -## Workflow - -1. **Identify parties and project scope** — GC/subcontractor/owner relationships, project type -2. **Scan for clause categories** — detect 100+ construction-specific clause types (indemnity, insurance, change orders, schedule, payment, retainage, flow-down, WAWF, Davis-Bacon, prevailing wage) -3. **Evaluate risk levels** — tag each clause as low/medium/high risk with plain-language rationale tied to source language -4. **Compare against standards** — check flagged clauses against client's preferred/fallback language -5. **Project-level risk aggregation** — summarize cumulative risk across the full contract, not just individual clauses -6. **Draft redline recommendations** — produce Word-compatible markup with revision language - -## Pitfalls - -- Don't confuse prime contract flow-down obligations with subcontract terms -- Davis-Bacon and prevailing wage requirements are mandatory — flag violations prominently -- Retainage and payment timing clauses interact — review together -- Construction-specific indemnity (broad-form vs. intermediate vs. limited) is a dealbreaker — don't miss it -- Schedule/delay clauses often hide liquidated damages exposure diff --git a/skills/legal/construction-contract-risk-summary/SKILL.md b/skills/legal/construction-contract-risk-summary/SKILL.md deleted file mode 100644 index bfb910f8..00000000 --- a/skills/legal/construction-contract-risk-summary/SKILL.md +++ /dev/null @@ -1,36 +0,0 @@ ---- -name: construction-contract-risk-summary -description: Generate jobsite-ready contract risk summaries for PMs and field leadership -tags: - practice: transactional - document: summary - mode: summarization ---- - -# Construction Contract Risk Summary - -Generate plain-language risk summaries from construction contracts for non-legal stakeholders. - -## Prerequisites - -- Completed contract risk analysis (clause-level findings) -- Project name, number, and location -- Client's role (GC, sub, owner, AE) -- Key financial terms (contract value, retainage %, payment timeline) - -## Workflow - -1. **Executive overview** — One-paragraph summary of the contract's key risks and opportunities -2. **Risk heat map** — Tabular view of risk areas by category (financial, safety, schedule, indemnity, insurance) -3. **Clause-level highlights** — Top 5–10 most critical clauses with plain-language explanations -4. **Action items** — Specific next steps: items requiring negotiation, items for insurance review, items to flag at pre-con -5. **Kickoff-ready format** — Structured output suitable for sharing at project kickoff meeting -6. **Go/no-go recommendation** — Based on cumulative risk profile - -## Pitfalls - -- Keep it jobsite-ready: PMs don't want legalese, they want to know what could blow up their schedule -- Financial risks (payment terms, retainage, LDs) are always priority one -- Don't overstate risks that are contractually mitigated (insurance, caps, waivers) -- Distinguish between GC risks vs. subcontractor risks — they're often opposite -- Always flag items that need to be discussed at pre-construction meeting diff --git a/skills/legal/construction-defect-analysis/SKILL.md b/skills/legal/construction-defect-analysis/SKILL.md deleted file mode 100644 index b494aa00..00000000 --- a/skills/legal/construction-defect-analysis/SKILL.md +++ /dev/null @@ -1,31 +0,0 @@ ---- -name: construction-defect-analysis -description: Analyze construction defects, classify by type, and generate defect reports for claim evaluation -tags: - - practice: regulatory - - document: analysis - - mode: analysis ---- - -# Construction Defect Analysis - -Document and analyze construction defects to support claims, change orders, and dispute resolution. - -## Prerequisites -- Photos or site inspection notes -- Original construction documents and specifications -- Warranty or guarantee terms -- Timeline of defect discovery - -## Workflow -1. **Classify**: Categorize defects by type (structural, envelope, MEP, finishes, code) -2. **Document**: Create a defect log with location, photos, description, and severity -3. **Trace**: Link each defect to the responsible trade, phase, or specification gap -4. **Cost Estimate**: Provide rough-order magnitude repair cost estimates -5. **Report**: Generate a formal defect report for claims or litigation support - -## Pitfalls -- Distinguish between latent and patent defects — statute of limitations differs -- Don't attribute defects to design without engineering analysis -- Preserve chain of custody for photographic evidence -- Check warranty periods — some defects may be covered, others are not diff --git a/skills/legal/construction-drawing-review/SKILL.md b/skills/legal/construction-drawing-review/SKILL.md deleted file mode 100644 index 55046902..00000000 --- a/skills/legal/construction-drawing-review/SKILL.md +++ /dev/null @@ -1,32 +0,0 @@ ---- -name: construction-drawing-review -description: Review construction drawings for discrepancies, code violations, and design conflicts before submission -tags: - - practice: regulatory - - document: analysis - - mode: analysis ---- - -# Construction Drawing Review - -Systematically review architectural and engineering drawings to identify discrepancies, code violations, and conflicts. - -## Prerequisites -- Full plan set (PDF drawings) -- Applicable building codes (ICC, NFPA, ADA, FHA, local amendments) -- Project type and jurisdiction -- Any client-specific requirements or constraints - -## Workflow -1. **Extract**: Parse drawings to identify building type, systems, and specifications -2. **Cross-Check**: Compare drawings across disciplines (architectural, structural, MEP) -3. **Code Verify**: Check compliance against applicable codes and local amendments -4. **Conflict Detect**: Flag spatial conflicts, dimension mismatches, and resolution gaps -5. **Report**: Generate a numbered issues list with location references and code citations -6. **Fix Suggest**: Provide AI-supported solutions for each flagged issue - -## Pitfalls -- Always verify which code edition applies to the jurisdiction and project date -- Check for local amendments — generic code checks miss these -- Don't confuse drawing discrepancies with design decisions the client prefers -- Version control matters — confirm you're reviewing the latest revision diff --git a/skills/legal/contract-leakage-analysis/SKILL.md b/skills/legal/contract-leakage-analysis/SKILL.md deleted file mode 100644 index 8ffbc431..00000000 --- a/skills/legal/contract-leakage-analysis/SKILL.md +++ /dev/null @@ -1,29 +0,0 @@ ---- -name: Contract Leakage Analysis -description: Identify revenue leakage from missed renewals, unclaimed rebates, auto-renewal at unfavorable terms, and margin erosion. Quantifies recoverable value and provides negotiation leverage data. -tags: - - practice_area: transactional - - document_type: analysis, checklist - - skill_mode: analysis ---- - -# Contract Leakage Analysis - -## Prerequisites -- Portfolio of executed contracts -- Financial data (invoices, payments, rebate records) -- Supplier/vendor spend data (if available) - -## Workflow -1. **Analyze contract portfolio** — scan all contracts for financial terms, pricing, and renewal language -2. **Identify leakage points** — flag missed auto-renewals, unclaimed rebates, expired discount terms, unfavorable rate escalations -3. **Quantify recoverable value** — calculate dollar amounts of leaked value by category and contract -4. **Benchmark terms** — compare against market rates and historical pricing to identify underperforming contracts -5. **Generate negotiation briefs** — produce data-backed positions for each contract renegotiation -6. **Prioritize by impact** — rank leakage sources by dollar value and recoverability - -## Pitfalls -- Leakage detection requires comparing contract terms to actual payments — cross-reference financial records -- Do not attribute all margin erosion to contract terms — market factors matter -- Historical pricing comparisons need context (volume changes, scope changes) -- Flag items that require legal review before negotiation — some leaks involve disputed terms diff --git a/skills/legal/contract-renewal-tracking/SKILL.md b/skills/legal/contract-renewal-tracking/SKILL.md deleted file mode 100644 index 119dcb2b..00000000 --- a/skills/legal/contract-renewal-tracking/SKILL.md +++ /dev/null @@ -1,29 +0,0 @@ ---- -name: Contract Renewal and Obligation Tracking -description: Track contract renewals, auto-renewal clauses, notice periods, and payment milestones proactively. Provides alerts on critical dates to prevent missed deadlines and operational chaos. -tags: - - practice_area: transactional - - document_type: checklist, policy - - skill_mode: drafting ---- - -# Contract Renewal and Obligation Tracking - -## Prerequisites -- Portfolio of executed contracts -- Known renewal preferences and negotiation timelines -- Internal approval workflows for renewals - -## Workflow -1. **Ingest contracts** — load contract portfolio for analysis -2. **Extract dates and deadlines** — identify renewal dates, auto-renewal triggers, notice periods, milestone dates -3. **Build Events Dashboard** — create centralized view of all contractual obligations and deadlines -4. **Set proactive alerts** — configure notifications for upcoming renewals, notice windows, and payment due dates -5. **Monitor compliance** — track whether obligations are being fulfilled on both sides -6. **Renewal preparation** — surface terms, pricing history, and negotiation options before renewal windows - -## Pitfalls -- Auto-renewal clauses can be buried in fine print — verify language carefully -- Notice periods vary widely; do not assume standard timeframes -- Track both sides' obligations — counterparties may also miss their commitments -- Maintain audit trail of all alerts and actions taken diff --git a/skills/legal/contract-review-triage/SKILL.md b/skills/legal/contract-review-triage/SKILL.md deleted file mode 100644 index 99c6ad34..00000000 --- a/skills/legal/contract-review-triage/SKILL.md +++ /dev/null @@ -1,38 +0,0 @@ -name: contract-review-triage -language: en -description: Prioritizes contracts for human review by ML-assisted risk assessment — scores contracts by complexity, risk exposure, and negotiation urgency. Enables legal service providers and law firms to triage high-volume document reviews efficiently. -tags: - - analysis - - checklist - - corporate ---- - -# ML-Assisted Contract Review Triage - -Scores and prioritizes contracts for legal review based on ML-identified risk factors, complexity indicators, and business urgency — enabling efficient triage of high-volume contract portfolios. - -## Prerequisites - -Before executing, collect: - -1. **Contract batch** — collection of contracts to triage (PDFs, DOCX, or text) -2. **Review criteria** — what matters most (risk, complexity, revenue impact, regulatory exposure) -3. **Risk thresholds** — what score ranges trigger urgent vs. routine vs. no-review -4. **Reviewer capacity** — how many contracts per reviewer per day determines prioritization - -## Workflow - -1. **Document ingestion** — batch load contracts with metadata (counterparty, type, execution date) -2. **ML risk scoring** — analyze each contract for risk signals: unusual terms, missing protections, asymmetric obligations -3. **Complexity scoring** — count clauses, cross-references, custom provisions vs. boilerplate ratio -4. **Business context** — weight scores by contract value, counterparty risk, regulatory environment -5. **Priority ranking** — sort contracts into triage tiers: critical review, standard review, light review, no-review -6. **Assignment** — route to appropriate reviewer based on specialty (commercial, employment, IP, etc.) -7. **Feedback loop** — capture reviewer corrections to improve ML model accuracy over time - -## Pitfalls - -- ML scores are probabilistic — low-risk scores can miss novel or jurisdiction-specific issues -- Scoring models need continuous calibration — feedback from reviewer corrections is essential -- High-volume contexts may require different thresholds than one-off reviews -- Custom contracts from sophisticated counterparties need senior reviewer assignment regardless of score diff --git a/skills/legal/contract-term-extraction/SKILL.md b/skills/legal/contract-term-extraction/SKILL.md deleted file mode 100644 index 766d634a..00000000 --- a/skills/legal/contract-term-extraction/SKILL.md +++ /dev/null @@ -1,28 +0,0 @@ ---- -name: Contract Term Extraction -description: Extract key terms, financial details, risks, and obligations from executed contracts. Provides instant Q&A over contract content with source-linked answers. Cuts admin time by up to 30%. -tags: - - practice_area: transactional - - document_type: agreement, checklist, analysis - - skill_mode: analysis ---- - -# Contract Term Extraction - -## Prerequisites -- Executed contract document (PDF, Word, or scan) -- Areas of interest (financial terms, obligations, deadlines, risks) - -## Workflow -1. **Upload contract** — ingest document in any common format -2. **Auto-extract key terms** — identify parties, dates, amounts, obligations, and special clauses -3. **Generate SnapShot** — produce structured extraction summary with risks and financial details -4. **Enable Q&A** — allow natural language queries over contract content with source-linked answers -5. **Cross-reference** — compare extracted terms against playbooks or historical agreements -6. **Export results** — deliver structured data, summaries, or full review report - -## Pitfalls -- Always verify extracted terms against source text — AI can miss nuanced language -- Financial amounts and dates require double-checking -- Ambiguous clauses should be flagged for human review -- Never assume a term is standard without verification diff --git a/skills/legal/covenant-fund-investment-review/SKILL.md b/skills/legal/covenant-fund-investment-review/SKILL.md deleted file mode 100644 index 0de64787..00000000 --- a/skills/legal/covenant-fund-investment-review/SKILL.md +++ /dev/null @@ -1,37 +0,0 @@ ---- -name: covenant-fund-investment-review -description: AI-powered legal review of fund investment documents including private equity, venture capital, and hedge fund formations -tags: - - practice: corporate, regulatory - - document: summary, analysis, memo - - mode: analysis, summarization ---- - -# Covenant Fund Investment Review - -AI-generated legal summaries, issue spotting, and commentary for alternative investment documents (VC, PE, hedge funds). - -## Prerequisites - -- Investment document (LPA, subscription agreement, side letter, NVCA form) -- Negotiation context (LP/GP relationship, market norms for fund vintage/size) -- Reference framework: standard NVCA terms, market benchmark for fee structures -- Client risk tolerance and carry waterfall preferences - -## Workflow - -1. Ingest investment document and classify document type (LPA, subscription, side letter, SAFE) -2. Extract key economic terms: management fees, carry %, hurdle rate, waterfall structure -3. Generate summary with plain-English commentary on each material term -4. Identify non-standard provisions and flag items requiring negotiation -5. Produce issues list ranked by severity: structural, economic, governance, reporting -6. Compare against historical portfolio terms for MFN analysis -7. Deliver formatted review within 48 hours: summary + commentary + issues list - -## Pitfalls - -- AI-generated summaries are for informational purposes; never substitute for licensed legal advice -- Fund structures vary widely (GFS vs. GFS-Plus, clawback mechanisms); ensure model is trained on the specific structure -- Carry waterfall nuances (European vs. American, deal-by vs. whole-of-fund) require careful analysis -- MFN provisions require comparison across ALL existing side letters — incomplete data yields false conclusions -- Regulatory changes (SEC, ESMA) may render standard terms outdated; always cross-check current rules diff --git a/skills/legal/covenant-mfns-election-analysis/SKILL.md b/skills/legal/covenant-mfns-election-analysis/SKILL.md deleted file mode 100644 index 2f381d11..00000000 --- a/skills/legal/covenant-mfns-election-analysis/SKILL.md +++ /dev/null @@ -1,37 +0,0 @@ ---- -name: covenant-mfns-election-analysis -description: Analyze Most Favored Nation elections in side letter provisions and compare terms across LP portfolio investments -tags: - - practice: corporate - - document: analysis, memo, summary - - mode: analysis, research ---- - -# Covenant MFN Election Analysis - -AI-powered analysis of MFN election rights in fund side letters, comparing terms across portfolio investments. - -## Prerequisites - -- LP's existing side letters with fee/carry terms -- New fund's proposed terms (or draft side letter) -- MFN clause text from LP's most favorable existing agreement -- Fund structure context (vintage, strategy, track record) - -## Workflow - -1. Extract MFN clause from LP's best existing side letter -2. Parse new fund's proposed terms for fee, carry, hurdles, and governance -3. Compare new terms against MFN benchmark across all comparable provisions -4. Identify where new terms are less favorable than existing MFN terms -5. Generate negotiation position paper with specific clause-level requests -6. Produce summary: which provisions trigger MFN, recommended asks, fallback positions -7. Auto-generate updated side letter draft incorporating MFN-entitled terms - -## Pitfalls - -- MFN clauses have narrow trigger conditions; verify the new fund qualifies under existing MFN text -- Not all terms are MFN-eligible (e.g., governance provisions may be excluded from MFN scope) -- Fund vintage and strategy matter: comparison must use peers with similar characteristics -- MFN rights may have sunset clauses or expiration dates; verify election windows are open -- LP relationships vary — some GPs resist MFN demands; prepare justification language tied to market norms diff --git a/skills/legal/covenant-nda-markup-triage/SKILL.md b/skills/legal/covenant-nda-markup-triage/SKILL.md deleted file mode 100644 index 3816b4ff..00000000 --- a/skills/legal/covenant-nda-markup-triage/SKILL.md +++ /dev/null @@ -1,36 +0,0 @@ ---- -name: covenant-nda-markup-triage -description: Streamline NDA markup and triage workflows with AI-assisted review for in-house legal teams and paralegals -tags: - - practice: corporate, transactional - - document: agreement, letter - - mode: analysis, drafting ---- - -# Covenant NDA Markup Triage - -AI-assisted NDA review and markup for high-volume deal flow and routine legal matters. - -## Prerequisites - -- Master NDA template (mutual or unilateral) -- Standard markup position paper with firm's negotiating stance -- Deal context: parties, purpose of disclosure, jurisdiction - -## Workflow - -1. Ingest counterparty's NDA draft and classify as mutual or unilateral -2. Auto-detect non-standard provisions: overbroad definitions, excessive term length, broad non-solicitation -3. Compare against firm's standard position on each clause -4. Generate markup with tracked changes and inline commentary explaining each change -5. Flag clauses requiring senior attorney approval vs. auto-approve routine items -6. Produce clean version and markup version with negotiation memo -7. Track NDA execution and set reminder for term expiration - -## Pitfalls - -- Overbroad definition of "Confidential Information" can inadvertently cover public knowledge; verify carve-outs -- Non-disclosure periods >5 years are non-market for most deal types; flag for negotiation -- Residual clauses favoring the receiving party create long-term competitive risks -- Governing law clauses shift litigation cost — always verify jurisdiction aligns with client's home -- NDA markups accumulate over time; maintain a living playbook of what's been conceded to avoid repeating trade-offs diff --git a/skills/legal/deepjudge-agile-workflows/SKILL.md b/skills/legal/deepjudge-agile-workflows/SKILL.md deleted file mode 100644 index 53bb6eb7..00000000 --- a/skills/legal/deepjudge-agile-workflows/SKILL.md +++ /dev/null @@ -1,24 +0,0 @@ ---- -name: deepjudge-agile-workflows -description: Design LLM-powered AI workflows for repetitive legal operations — multi-step research, document synthesis, and cross-referencing across matter files. -tags: [legal-ops, regulatory] -skill_mode: [drafting, analysis] ---- - -# Legal AI Workflow Design - -## Prerequisites -- Catalog of repetitive legal tasks in the practice area -- Available data sources (documents, APIs, internal systems) - -## Workflow -1. Inventory the repetitive task: inputs, decision points, outputs, stakeholders -2. Decompose into discrete steps; identify which need LLM reasoning vs. rule-based logic -3. Design the workflow graph: trigger → retrieve → reason → act → review -4. Specify guardrails: LLM-agnostic reasoning, data source constraints, audit logging -5. Draft the workflow specification with fallback human review points - -## Pitfalls -- Never fully automate without a human-in-the-loop checkpoint on consequential outputs -- Keep workflows LLM-agnostic; don't hardcode model-specific behaviors -- Log every retrieval and action for auditability diff --git a/skills/legal/deepjudge-knowledge-search/SKILL.md b/skills/legal/deepjudge-knowledge-search/SKILL.md deleted file mode 100644 index d88c1ebc..00000000 --- a/skills/legal/deepjudge-knowledge-search/SKILL.md +++ /dev/null @@ -1,24 +0,0 @@ ---- -name: deepjudge-knowledge-search -description: Draft enterprise legal knowledge search queries and results analysis to surface prior cases, negotiated terms, and matter-specific insights from institutional knowledge bases. -tags: [legal-research, litigation, corporate] -skill_mode: [research, analysis] ---- - -# DeepJudge-Style Institutional Knowledge Search - -## Prerequisites -- Access to firm/in-house knowledge base or document repository -- Understanding of the legal matter's factual context - -## Workflow -1. Map the matter's key factual predicates, parties, and legal issues -2. Formulate entity-linked search queries combining prior matter details with legal topics -3. Execute multi-source retrieval across documents, negotiations, and case files -4. Score and rank results by similarity to current matter's factual matrix -5. Synthesize findings into a research memo highlighting analogous situations, outcomes, and negotiated positions - -## Pitfalls -- Don't retrieve without a clear hypothesis — vague queries return noise -- Prior matter info must be anonymized before search; preserve confidentiality -- Cross-reference results against jurisdiction and timeline constraints diff --git a/skills/legal/deepjudge-legal-intelligence-gathering/SKILL.md b/skills/legal/deepjudge-legal-intelligence-gathering/SKILL.md deleted file mode 100644 index 40ff2cf4..00000000 --- a/skills/legal/deepjudge-legal-intelligence-gathering/SKILL.md +++ /dev/null @@ -1,24 +0,0 @@ ---- -name: deepjudge-legal-intelligence-gathering -description: Create matter overviews and timelines by aggregating events, documents, and communications across a client file or folder structure. -tags: [litigation, corporate] -skill_mode: [summarization, analysis] ---- - -# Matter Timeline & Overview Generation - -## Prerequisites -- Access to matter file structure (documents, emails, communications) -- Understanding of the matter's scope and relevant time period - -## Workflow -1. Ingest all documents in the matter file; extract key entities, dates, and events -2. Build an event timeline chronologically, deduplicating overlapping records -3. Classify events by category: filings, negotiations, communications, decisions -4. Generate a narrative overview summarizing the matter's trajectory and current posture -5. Highlight gaps in the record where information may be missing or delayed - -## Pitfalls -- Don't invent events from ambiguous records — flag uncertainty explicitly -- Maintain chain-of-custody for all source documents cited in the overview -- Separate factual summaries from legal analysis to avoid overreach diff --git a/skills/legal/deposition-interrogation/SKILL.md b/skills/legal/deposition-interrogation/SKILL.md deleted file mode 100644 index 2a59cf41..00000000 --- a/skills/legal/deposition-interrogation/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ ---- -name: deposition-interrogation -description: AI-assisted interrogation of deposition testimony using natural language questions -tags: - practice: litigation - document: analysis - mode: research ---- - -# Deposition Interrogation - -Ask natural language questions about deposition testimony and get cited, verbatim answers from the record. - -## Prerequisites - -- Uploaded deposition transcript or video recording -- Witness name and role in the case -- Relevant document set for cross-reference - -## Workflow - -1. **Load testimony record** — Import transcript or video with AI indexing -2. **Ask questions naturally** — Pose questions in plain language ("What did the witness say about the timeline?") -3. **Get cited answers** — Receive verbatim quotes with video timestamp and transcript page/line references -4. **Follow-up追问** — Drill deeper into answers with follow-up questions -5. **Memory consistency check** — Compare witness statements across multiple deposition sessions -6. **Gap identification** — Flag critical topics the witness avoided or contradicted themselves on - -## Pitfalls - -- AI can hallucinate — always verify answers against the actual record -- Don't ask compound questions — break them into single-focus inquiries -- Witness "I don't recall" answers need context — check what they knew when -- Timeline questions are high-value but easy to mess up — verify chronology carefully -- The AI's "understanding" of answers is not legal analysis — interpret results yourself diff --git a/skills/legal/deposition-searchable-record/SKILL.md b/skills/legal/deposition-searchable-record/SKILL.md deleted file mode 100644 index 1b05d4ac..00000000 --- a/skills/legal/deposition-searchable-record/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ ---- -name: deposition-searchable-record -description: Build a searchable deposition record for rapid discovery of key testimony moments -tags: - practice: litigation - document: analysis - mode: research ---- - -# Searchable Deposition Record - -Create a comprehensive, searchable deposition record for rapid discovery of key testimony. - -## Prerequisites - -- Deposition video or transcript files -- Case keyword list and witness key topics -- Prior related depositions for comparison - -## Workflow - -1. **Ingest deposition** — Upload video; system produces timecoded transcript with AI indexing -2. **Index key topics** — AI automatically tags testimony by topic (liability, damages, knowledge, intent, etc.) -3. **Full-text search** — Search across entire deposition for any term, phrase, or concept -4. **Topic browsing** — Navigate deposition by AI-generated topic clusters -5. **Compare against prior depositions** — Cross-reference with other witness testimony for patterns -6. **Export for trial prep** — Generate printable excerpts with timestamps for exhibit preparation - -## Pitfalls - -- AI topic tagging is assistive — verify that topics match the case theory -- Search results need context — a keyword match doesn't mean the answer supports your theory -- Video timestamps may drift between platforms — always verify critical references -- Search doesn't replace reading the full transcript — use as a discovery tool, not a substitute -- Redacted or sealed portions must be clearly marked to avoid inadvertent use diff --git a/skills/legal/deposition-video-analysis/SKILL.md b/skills/legal/deposition-video-analysis/SKILL.md deleted file mode 100644 index 64b0c43d..00000000 --- a/skills/legal/deposition-video-analysis/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ ---- -name: deposition-video-analysis -description: Analyze deposition video recordings to extract key testimony, inconsistencies, and impeachment material -tags: - practice: litigation - document: summary - mode: analysis ---- - -# Deposition Video Analysis - -Video-first deposition analysis that treats testimony as a searchable, reviewable, and reusable asset. - -## Prerequisites - -- Deposition video recording (Zoom-hosted or uploaded footage) -- Case theory and key testimony targets -- Prior deposition transcripts for comparison - -## Workflow - -1. **Generate transcript** — Produce ASR-powered rough transcript with timecoded video references -2. **Video-first navigation** — Navigate testimony by video timestamp rather than page/line -3. **Searchable keywords** — Search across testimony for specific words, phrases, and concepts -4. **AI interrogation** — Ask natural language questions about witness testimony and get cited answers -5. **Inconsistency detection** — Compare current deposition against prior depositions and documents -6. **Impeachment builder** — Compile key quotes with video timestamps for trial use - -## Pitfalls - -- ASR accuracy is high but not perfect — flag sections needing human verification -- Video-first doesn't replace page/line citation — keep both formats for court filings -- Timestamps must be exact — a wrong reference is a credibility killer -- Don't rely solely on AI summarization for critical testimony — verify with raw video -- Cross-reference with prior transcripts — inconsistencies between depositions are powerful impeachment tools diff --git a/skills/legal/dexterity-deal-intelligence-dashboard/SKILL.md b/skills/legal/dexterity-deal-intelligence-dashboard/SKILL.md deleted file mode 100644 index c5fc656c..00000000 --- a/skills/legal/dexterity-deal-intelligence-dashboard/SKILL.md +++ /dev/null @@ -1,37 +0,0 @@ ---- -name: dexterity-deal-intelligence-dashboard -description: Organize deal negotiation intelligence in real-time dashboards, aggregating data from email threads and document versions -tags: - - practice: corporate, transactional - - document: summary, checklist - - mode: analysis, research ---- - -# Dexterity Deal Intelligence Dashboard - -Aggregate and organize deal negotiation data from scattered email threads and document versions into a unified real-time dashboard. - -## Prerequisites - -- Email integration (Microsoft Outlook or Gmail) -- Document storage with version history -- Deal structure document mapping key terms to file locations -- Negotiation timeline tracking - -## Workflow - -1. Connect email and document systems to the deal workspace -2. Auto-classify incoming communications: draft version received, markup sent, term confirmation, internal note -3. Map key terms from each document version to a live comparison table -4. Track negotiation status per clause: open, countered, agreed, disputed -5. Generate deal health dashboard: version age, response time, bottleneck clauses, remaining items -6. Alert on stale items (>48h no response) and escalate to responsible party -7. Export consolidated term sheet from latest document versions - -## Pitfalls - -- Email parsing misses non-email communications (Slack, phone calls, meeting notes); document all outside communications -- Version conflicts: two parties may submit markups simultaneously; detect and flag concurrent edits -- Dashboard data is only as good as the source; manual entry of verbal agreements is essential -- High-volume email threads can obscure critical term changes; prioritize clause-level tracking over full-document review -- Client-facing dashboards must sanitize internal notes and flagged items before external sharing diff --git a/skills/legal/dexterity-secure-contract-negotiation/SKILL.md b/skills/legal/dexterity-secure-contract-negotiation/SKILL.md deleted file mode 100644 index 1c5c36d0..00000000 --- a/skills/legal/dexterity-secure-contract-negotiation/SKILL.md +++ /dev/null @@ -1,37 +0,0 @@ ---- -name: dexterity-secure-contract-negotiation -description: Manage secure digital contract negotiations with real-time version tracking and confidential document handling -tags: - - practice: corporate, transactional - - document: agreement, letter - - mode: drafting, analysis ---- - -# Dexterity Secure Contract Negotiation - -Secure digital negotiation platform for managing contract drafts, version control, and confidential deal communications. - -## Prerequisites - -- Negotiation-ready document template with defined clause library -- Secure document sharing link (not email attachment) -- Clause negotiation playbook with fallback positions -- Role-based access controls for deal team members - -## Workflow - -1. Initiate negotiation by generating secure document link with version history -2. Distribute initial draft; all parties work in the same live environment -3. Track changes in real-time with per-clause attribution and timestamp -4. Surface clause-level negotiation status: proposed, countered, agreed, dropped -5. Flag conflicting versions before consolidation -6. Generate clean executed version from final negotiated state -7. Archive complete negotiation history with audit trail for compliance - -## Pitfalls - -- Secure platforms may exclude external counsel not on the platform; plan fallback for third-party access -- Version merge conflicts require manual resolution; don't assume auto-merge is always correct -- Clause-level tracking requires disciplined use — ad-hoc comments outside the system break the audit trail -- Redlining conventions differ across jurisdictions; establish shared markup standards at negotiation start -- Confidential information visible to all platform users; verify everyone on the deal workspace has a bona fide need diff --git a/skills/legal/dexterity-version-control-deals/SKILL.md b/skills/legal/dexterity-version-control-deals/SKILL.md deleted file mode 100644 index 61c3f37a..00000000 --- a/skills/legal/dexterity-version-control-deals/SKILL.md +++ /dev/null @@ -1,37 +0,0 @@ ---- -name: dexterity-version-control-deals -description: Manage document version control for deal negotiations with automated comparison and consolidated term extraction -tags: - - practice: corporate - - document: agreement, summary - - mode: analysis, drafting ---- - -# Dexterity Deal Version Control - -Automated document version comparison and term consolidation for multi-round deal negotiations. - -## Prerequisites - -- Source document with structured clause numbering or headings -- Clause-by-clause negotiation tracker -- Historical document versions stored in accessible location -- Change summary template for each negotiation round - -## Workflow - -1. Import all document versions for the negotiation thread -2. Run automated comparison engine across versions to identify clause-level changes -3. Generate version delta summary: added, removed, modified, unchanged per clause -4. Extract current negotiated state of each term from the latest version -5. Produce consolidated term sheet reflecting the final agreed position on every clause -6. Compare consolidated terms against original deal expectations and identify gaps -7. Deliver version-control report to deal team with clear recommendations on remaining open items - -## Pitfalls - -- Auto-comparison may miss subtle wording changes that shift legal meaning; always do manual clause review -- Structured clauses with cross-references break comparison engines; verify cross-references after merge -- Redline-only tracking loses negotiated compromises hidden in markup; ensure clean versions are also ingested -- Term extraction assumes clause headings match across versions; rename anomalies break mapping -- Don't rely on version count as deal health indicator — 20 versions with no term movement means stale negotiations diff --git a/skills/legal/dyspute-ai-mediation-assistant/SKILL.md b/skills/legal/dyspute-ai-mediation-assistant/SKILL.md deleted file mode 100644 index 42a5bfb1..00000000 --- a/skills/legal/dyspute-ai-mediation-assistant/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ ---- -name: ai-mediation-assistant -description: Provide AI-assisted mediation guidance with neutral analysis of positions, proposals, and settlement pathways for dispute resolution -tags: - - practice_area: litigation - - document_type: memo, summary - - skill_mode: analysis ---- - -# AI-Mediation Assistant - -Neutral AI assistant for structured mediation, providing position analysis, proposal evaluation, and settlement pathway mapping. - -## Prerequisites - -- Case brief for each party -- Known positions and demands from both sides -- Mediation timeline and venue constraints - -## Workflow - -1. Ingest both parties' case summaries and positions -2. Identify overlapping interests and fundamental disagreements -3. Generate neutral position summaries (same for both sides, no bias) -4. Evaluate proposed settlement ranges and identify the negotiation gap -5. Suggest proposal adjustments that could bridge the gap -6. Track proposal rounds, voting outcomes, and convergence metrics -7. Generate settlement agreement draft when consensus is reached - -## Pitfalls - -- Must remain strictly neutral — never favor one party's position -- Proposal suggestions should be framed as options, not directives -- Settlement amounts must include confidence intervals, not point estimates -- Document all proposal rounds for the mediation record diff --git a/skills/legal/dyspute-negotiation-proposal-generation/SKILL.md b/skills/legal/dyspute-negotiation-proposal-generation/SKILL.md deleted file mode 100644 index 647e5db8..00000000 --- a/skills/legal/dyspute-negotiation-proposal-generation/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ ---- -name: ai-negotiation-proposal-generation -description: Generate AI-assisted negotiation proposals with voting mechanisms, direct offers, and settlement agreement automation -tags: - - practice_area: transactional - - document_type: agreement, letter - - skill_mode: drafting ---- - -# AI Negotiation Proposal Generator - -Generate structured negotiation proposals, support voting between AI-generated and human-crafted offers, and auto-generate settlement agreements. - -## Prerequisites - -- Current offer history and voting results -- Known bottom-lines and constraints for the proposing party -- Settlement framework (monetary, non-monetary, or hybrid) - -## Workflow - -1. Review all prior proposals and voting results from negotiation rounds -2. Analyze the convergence pattern: is the gap narrowing, widening, or stagnant? -3. Generate AI-assisted proposals that optimize for the proposer's constraints -4. Present voting options: AI-generated proposal vs. direct human-crafted offer -5. Execute voting round and record results -6. When proposals converge, auto-generate the settlement agreement -7. Process settlement payment through integrated payment flow - -## Pitfalls - -- AI proposals must respect party-imposed constraints (budget caps, non-negotiable terms) -- Voting results must be documented per-round with rationale -- Settlement agreement generation must include all negotiated terms, not assumed defaults -- Payment processing integration must be verified before generating settlement docs diff --git a/skills/legal/ecommerce-infringement-analysis/SKILL.md b/skills/legal/ecommerce-infringement-analysis/SKILL.md deleted file mode 100644 index 772739d1..00000000 --- a/skills/legal/ecommerce-infringement-analysis/SKILL.md +++ /dev/null @@ -1,38 +0,0 @@ -name: ecommerce-infringement-analysis -language: en -description: Analyzes suspected ecommerce infringements to classify severity — distinguishes direct counterfeiting, unauthorized reselling, image copyright theft, and trademark confusion. Produces actionable enforcement recommendations with platform-specific next steps and estimated removal success probability. -tags: - - analysis - - agreement - - regulatory ---- - -# Ecommerce Infringement Classification & Analysis - -Classifies and prioritizes ecommerce IP infringements by type and severity, then recommends platform-appropriate enforcement actions with success probability estimates. - -## Prerequisites - -Before executing, collect: - -1. **Infringing listing(s)** — URL(s) to suspected infringing product pages -2. **Owner's IP rights** — trademark registrations, copyright certificates, design patents -3. **Product reference data** — authorized product images, descriptions, MSRP -4. **Authorization proof** — brand registry status, letter of authorization status - -## Workflow - -1. **Listing analysis** — scrape infringing page: images, text, price, seller info, marketplace -2. **IP match scoring** — compare against registered marks/images using text and visual matching -3. **Infringement classification** — categorize as: direct counterfeiting, unauthorized resale, image theft, trademark confusion, or design patent violation -4. **Severity assessment** — rank by volume of infringement, price undercutting, brand damage risk, and seller history -5. **Enforcement recommendation** — select optimal action: DMCA notice, LLP/Brand Registry report, VeRO report, or cease-and-desist -6. **Success probability** — estimate removal likelihood based on marketplace, IP strength, and seller type -7. **Evidence package** — compile screenshots, IP matches, timestamps for enforcement submission - -## Pitfalls - -- Counterfeiters blur/mutate images to evade detection — require human verification -- Unauthorized resellers may have legitimate sourcing — distinguish from counterfeiting -- Platform-specific reporting forms have different evidentiary standards -- Some marketplaces (Temu, Shein, TikTok Shop) have inconsistent enforcement responsiveness diff --git a/skills/legal/enforce-autonomous-takedown/SKILL.md b/skills/legal/enforce-autonomous-takedown/SKILL.md deleted file mode 100644 index 58b2a758..00000000 --- a/skills/legal/enforce-autonomous-takedown/SKILL.md +++ /dev/null @@ -1,38 +0,0 @@ -name: enforce-autonomous-takedown -language: en -description: Automates end-to-end IP enforcement — from listing detection to marketplace takedown submission. Triggers on confirmed infringement, gathers evidence, formats DMCA/LLP notices, and manages removal workflows across 100+ platforms including Amazon, Shopify, and TikTok. -tags: - - analysis - - agreement - - regulatory ---- - -# Autonomous IP Enforcement — Takedown Automation - -Automates the full lifecycle of IP enforcement: detects infringements across marketplaces, validates ownership claims, prepares takedown notices, and manages removal workflows end-to-end. - -## Prerequisites - -Before executing, collect: - -1. **IP ownership proof** — trademark registration numbers, copyright certificates, or design patents -2. **Brand/product catalogue** — product images, SKUs, authorized sellers list -3. **Infringement targets** — specific URLs, marketplace names, or brand names being violated -4. **Authorization** — signed letter of authorization for on-platform takedowns - -## Workflow - -1. **Infringement intake** — receive suspected listing URLs or marketplace names -2. **Ownership validation** — verify IP rights against provided certificates/registrations -3. **Evidence assembly** — scrape infringing content, capture screenshots, record timestamps -4. **Notice preparation** — generate platform-appropriate notice (DMCA, LLP, Trademark counterfeiting) -5. **Takedown submission** — submit through marketplace enforcement channels (Seller Remedy, Brand Registry, etc.) -6. **Status tracking** — monitor removal progress, handle appeals/re submissions, log outcomes - -## Pitfalls - -- Platform-specific notice requirements vary (Amazon LLP vs DMCA vs eBay VeRO differ) -- Resubmissions common for gateway marketplace sellers (Alibaba, DHGate) -- Counterfeiters evolve tactics — static detection fails, need adaptive scanning -- Letter of authorization needed for platform-mediated removals -- Localized listings reappear from overseas factories diff --git a/skills/legal/expert-witness-matching/SKILL.md b/skills/legal/expert-witness-matching/SKILL.md deleted file mode 100644 index 424f3205..00000000 --- a/skills/legal/expert-witness-matching/SKILL.md +++ /dev/null @@ -1,31 +0,0 @@ ---- -name: Expert Witness Matching -description: Identify and recommend qualified board-certified medical expert witnesses for malpractice, PI, and pharmaceutical cases. Matches expertise, geography, availability, and past testimony record. -tags: - - practice_area: litigation - - document_type: analysis, summary - - skill_mode: research ---- - -# Expert Witness Matching - -## Prerequisites -- Case specialty and sub-specialty needed -- Jurisdiction and venue -- Case type (malpractice, PI, pharmaceutical, device) -- Geographic preference for expert location -- Budget constraints (optional) - -## Workflow -1. **Define requirements** — specialty, sub-specialty, jurisdiction, case type -2. **Search expert network** — filter by board certification, clinical practice, and credentials -3. **Verify qualifications** — confirm active licensure, board certification, and peer reviews -4. **Check testimony history** — review past depositions, trials, Daubert challenges -5. **Match logistics** — verify availability, geographic proximity, and compensation expectations -6. **Present ranked options** — deliver top matches with credentials, testimony record, and availability - -## Pitfalls -- Verify board certification is current and in good standing -- Watch for experts with high disqualification rates or failed Daubert challenges -- Check for conflicts of interest (prior relationships with parties or counsel) -- Avoid experts who testify primarily for one side exclusively diff --git a/skills/legal/firstdrafts-case-summary/SKILL.md b/skills/legal/firstdrafts-case-summary/SKILL.md deleted file mode 100644 index ff627ecf..00000000 --- a/skills/legal/firstdrafts-case-summary/SKILL.md +++ /dev/null @@ -1,34 +0,0 @@ ---- -name: firstdrafts-case-summary -tags: - practice: litigation - document: summary - mode: summarization ---- - -# Comprehensive Case Summary - -## Purpose -Generate a thorough case summary from uploaded pleadings, discovery documents, and briefs. Identifies key facts, chronology, parties, claims, defenses, and litigation posture — useful for trial prep, client updates, or handoffs. - -## Prerequisites -- All uploaded pleadings, complaints, answers, discovery responses -- Key dates and procedural milestones -- Party roles (plaintiff, defendant, third-party, intervenor) - -## Workflow -1. Build a chronological timeline from all documents: filing dates, service dates, hearing dates, deadlines -2. Identify and list each party with their role and claimed positions -3. Summarize each claim/cause of action with its factual basis and supporting evidence -4. Summarize each defense and affirmative defense with supporting facts -5. Extract discovery status: what has been produced, what is outstanding, motion status -6. Note the current litigation posture: dispositive motions pending, discovery complete, trial date set -7. Flag inconsistencies between parties' versions of material facts -8. Produce a one-page executive overview followed by detailed sections - -## Pitfalls -- Do not assume facts not contained in the uploaded documents -- Flag any timeline gaps where dates are missing -- Distinguish between allegation and proven fact clearly -- Note when different parties' chronologies conflict -- If document sources are unlabeled or unnamed, flag them for attorney cleanup diff --git a/skills/legal/firstdrafts-judge-ruling-simulator/SKILL.md b/skills/legal/firstdrafts-judge-ruling-simulator/SKILL.md deleted file mode 100644 index a1411ede..00000000 --- a/skills/legal/firstdrafts-judge-ruling-simulator/SKILL.md +++ /dev/null @@ -1,36 +0,0 @@ ---- -name: judge-ruling-simulator -tags: - practice: litigation - document: analysis - mode: analysis ---- - -# Judge Ruling Simulator - -## Purpose -Simulate how a judge might rule on a pending motion or dispute. Analyzes the argument positions from both sides, applies the governing legal standard, and predicts outcomes with reasoning. Used for pre-argument preparation and settlement valuation. - -## Prerequisites -- Full briefing on the motion in question (movant's memorandum and opposing brief) -- The specific judge or court (for behavioral context) -- Governing standard of review -- Material facts in dispute vs. undisputed - -## Workflow -1. Identify the precise legal question the judge must decide -2. Separate undisputed facts from disputed facts -3. Analyze the movant's strongest arguments and their legal support -4. Analyze the opponent's strongest counterarguments and their legal support -5. Apply the governing standard of review to weigh each side -6. Produce a predicted ruling for each argument: (a) granted, (b) denied, (c) partially granted -7. For each predicted ruling, provide the judge's likely reasoning -8. Identify the weakest link in your side's argument — the argument most likely to cause denial -9. Suggest supplementary authority or factual development to shore up weak points - -## Pitfalls -- This is a probabilistic simulation, not legal advice — clearly label predictions as estimates -- Avoid overconfidence; real judges have individual tendencies this cannot fully capture -- Do not assume facts not in the briefing record -- Flag when the legal question is genuinely unsettled and predictions are unreliable -- Remember: this tool does not replace attorney judgment on settlement strategy diff --git a/skills/legal/firstdrafts-litigation-motion/SKILL.md b/skills/legal/firstdrafts-litigation-motion/SKILL.md deleted file mode 100644 index fde0d060..00000000 --- a/skills/legal/firstdrafts-litigation-motion/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ ---- -name: firstdrafts-litigation-motion -tags: - practice: litigation - document: motion - mode: drafting ---- - -# Litigation Motion Drafting - -## Purpose -Draft a first-draft litigation motion (e.g., motion to dismiss, motion for summary judgment, motion in limine) based on case-specific facts and the attorney's prior writing style. Produces a persuasive, jurisdiction-appropriate motion ready for attorney review and refinement. - -## Prerequisites -- Statement of case facts (from brief, client memo, or case summary) -- Type of motion requested (MSJ, TDM, in limine, etc.) -- Governing standard of review for the motion type -- Prior similar motions or filings by the same attorney (style reference) -- Jurisdiction-specific procedural rules (FRCP / state equivalent) - -## Workflow -1. Restate the case background concisely using provided facts -2. State the legal standard of review for this motion type -3. For each ground for relief: (a) state the ground, (b) apply facts to law, (c) cite authority -4. Address anticipated opposing arguments and pre-rebuttal them -5. Draft the requested relief in precise, court-ready language -6. Format according to local court rules (numbered paragraphs, caption style) -7. Include a proposed order for the judge to sign - -## Pitfalls -- The AI does not have live internet access to verify current case law — flag all citations for attorney verification -- Do not fabricate procedural history; only use facts provided -- Match the attorney's voice from prior filings — not generic or overly formal language -- Ensure the motion's requested relief is available under the stated legal standard -- Never include attorney-client privileged material in the public filing section diff --git a/skills/legal/flipthrough-contract-proposal-analysis/SKILL.md b/skills/legal/flipthrough-contract-proposal-analysis/SKILL.md deleted file mode 100644 index 5ce379b5..00000000 --- a/skills/legal/flipthrough-contract-proposal-analysis/SKILL.md +++ /dev/null @@ -1,34 +0,0 @@ ---- -name: flipthrough-contract-proposal-analysis -tags: - practice: corporate - document: analysis - mode: analysis ---- - -# Contract and Proposal Analysis - -## Purpose -Analyze vendor proposals, MSAs, SOWs, SLAs, and licensing agreements. Identify key terms, risk exposures, liabilities, and negotiation leverage points. Output is a structured analysis with scoring and recommendations. - -## Prerequisites -- The contract or proposal document (PDF, Word, or text) -- Buyer's standard terms and playbooks (if available) -- Procurement category (IT, services, cloud, etc.) -- Deal context: competitive or sole-source - -## Workflow -1. Classify the document type (MSA, SOW, SLA, license, proposal, NDA) and extract the header terms: parties, effective date, term, governing law -2. Extract and analyze key clauses: indemnification, liability caps, IP ownership, data protection, termination, renewal, pricing, SLA/service credits -3. Score each clause against buyer's standard terms: (a) matches standard, (b)偏离 with rationale, (c) material deviation -4. Identify hidden risks: auto-renewal traps, unilateral change rights, broad IP grants, uncapped liability, onerous service credits -5. Rank issues by negotiation priority: (1) must-fix, (2) should-negotiate, (3) nice-to-have -6. Provide recommended alternative language for each material deviation -7. Produce a one-page deal summary with total risk score and negotiation strategy - -## Pitfalls -- Do not assume vendor documents are in good faith — flag all asymmetrical risk allocations -- Pricing terms in proposals are often placeholder — flag for commercial verification -- Service-level commitments in proposals may not appear in the final contract — flag for negotiation -- Beware of cross-references to other documents that change the terms materially -- Distinguish between the proposal (marketing document) and the actual contract (legal obligation) diff --git a/skills/legal/flipthrough-negotiation-briefing/SKILL.md b/skills/legal/flipthrough-negotiation-briefing/SKILL.md deleted file mode 100644 index fb327a8d..00000000 --- a/skills/legal/flipthrough-negotiation-briefing/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ ---- -name: flipthrough-negotiation-briefing -tags: - practice: corporate - document: letter - mode: drafting ---- - -# Contract Negotiation Briefing - -## Purpose -Generate a negotiation briefing document for internal teams before contract discussions with vendors. Includes deal context, priority items, tradeables, talk tracks, and fallback positions. - -## Prerequisites -- Draft contract or vendor proposal under negotiation -- Internal stakeholder positions (legal, procurement, security, finance) -- Known vendor constraints (budget, timeline, competitive landscape) -- Your organization's must-haves and tradeables list - -## Workflow -1. Header: document name, counterparty, deal value, key stakeholders, deadline -2. Situation summary: what is being negotiated and why it matters -3. Priority items: list each issue in order of importance with internal position -4. Talk tracks: for each priority, draft 2-3 opening statements and objection responses -5. Tradeables matrix: map items you can concede against items the vendor values -6. BATNA analysis: best alternative to negotiated agreement and walk-away threshold -7. Real-time markup guidance: highlight clauses where you should propose alternative language during live negotiation -8. Post-negotiation checklist: items to confirm, document, or escalate after the call - -## Pitfalls -- Keep talk tracks concise — 2 sentences max per objection response -- Never reveal your internal walk-away number in the briefing unless authorized -- Distinguish between firm positions and negotiable ones clearly -- Flag when a requested change is standard market practice vs. unique demand -- Update the briefing when new information emerges during negotiation diff --git a/skills/legal/flipthrough-vendor-risk-assessment/SKILL.md b/skills/legal/flipthrough-vendor-risk-assessment/SKILL.md deleted file mode 100644 index 8b84ba4c..00000000 --- a/skills/legal/flipthrough-vendor-risk-assessment/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ ---- -name: flipthrough-vendor-risk-assessment -tags: - practice: corporate - document: checklist - mode: analysis ---- - -# Vendor Contract Risk Assessment - -## Purpose -Systematically assess the risk profile of a vendor contract. Covers legal, financial, operational, and compliance dimensions. Produces a risk report with scoring and remediation recommendations. - -## Prerequisites -- Draft or executed vendor contract -- Vendor due diligence report (financial health, security posture, certifications) -- Internal risk framework categories (legal, financial, operational, compliance, reputational) -- Business criticality of the vendor relationship (tier 1/2/3) - -## Workflow -1. Legal risk: review indemnity, liability, IP, termination, dispute resolution clauses for risk -2. Financial risk: assess payment terms, penalty structures, termination costs, price escalation -3. Operational risk: evaluate SLA guarantees, service credits, data availability, business continuity -4. Compliance risk: check data protection, privacy, industry regulation, export controls, sanctions -5. Reputational risk: assess vendor's public record, litigation history, ESG posture -6. Score each category 1-5 (1=low risk, 5=critical risk) -7. For each score of 3+: draft a specific remediation recommendation -8. Produce overall risk score and go/no-go recommendation with conditions - -## Pitfalls -- A low score on one dimension does not offset a critical score on another — report highest risk prominently -- Do not rely solely on the vendor's self-reported data — flag items that need independent verification -- Context matters: a tier 3 vendor's operational risk is less critical than tier 1's -- Remember that risk assessments are snapshots — note the assessment date -- Never declare a vendor "low risk" without addressing compliance and data protection specifically diff --git a/skills/legal/genesis-case-evaluation-drafting/SKILL.md b/skills/legal/genesis-case-evaluation-drafting/SKILL.md deleted file mode 100644 index 064700dd..00000000 --- a/skills/legal/genesis-case-evaluation-drafting/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ -# Med-Mal & Mass Tort Case Evaluation Report Drafting - -## Prerequisites -- Patient medical records (all relevant encounter documentation) -- Relevant tort jurisdiction and statute of limitations -- Firm's case acceptance criteria and screening questionnaires - -## Workflow -1. Build a case chronology from all uploaded records: - - Index every clinical event by date and provider - - Identify the precipitating incident and relevant treatment timeline - - Flag gaps in the record (missing notes, unavailable records) -2. Apply the firm's case evaluation criteria: - - Duty of care — identify all treating providers and their standards - - Breach — compare treatment against applicable standard of care - - Causation — establish the causal link between breach and injury - - Damages — quantify economic and non-economic harm from records -3. Generate a case evaluation report: - - Factual narrative with date-stamped chronology - - Legal elements analysis with record citations for each element - - Strength assessment (strong/moderate/weak) with supporting rationale - - Recommended next steps (expert consultation, discovery priorities) -4. Include citation tracking — every factual assertion must reference a record page - -## Pitfalls -- Statute of limitations varies by jurisdiction — verify the clock starts at discovery -- Don't assume causation from temporal proximity — establish proximate cause -- Missing records can defeat a case — note omissions and subpoena needs -- Damages must be documented in the medical record — unreported pain isn't measurable -- Multi-plaintiff cases need individualized analysis — no blanket evaluations - -## Tags -practice_area: litigation -document_type: summary, analysis -skill_mode: analysis diff --git a/skills/legal/genesis-medical-chronology/SKILL.md b/skills/legal/genesis-medical-chronology/SKILL.md deleted file mode 100644 index 7774c972..00000000 --- a/skills/legal/genesis-medical-chronology/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ -# Automated Medical Record Chronology Generation - -## Prerequisites -- Complete set of client medical records (PDFs, scanned documents) -- Scope definition (full history vs. date range, specific providers) - -## Workflow -1. Ingest all records and index by document type, date, and authoring provider -2. Extract clinical events from each document: - - Visits/admissions and discharge summaries - - Physician and nursing notes, progress reports - - Diagnostic orders and results (labs, imaging, EKG) - - Medication administrations and changes - - Surgical/procedure notes and operative reports -3. Compile a master timeline: - - Sort all events chronologically - - Identify key milestones (diagnosis, treatment start, adverse events, discharge) - - Flag inconsistencies (date conflicts, missing entries, duplicate records) -4. Generate the chronology document: - - Date | Time | Provider | Document Type | Event Description | Source Citation - - Color-code by severity/clinical significance - - Include a summary of events with pattern indicators -5. Output a court-ready chronology with page citations for evidentiary use - -## Pitfalls -- Scanned documents may have incorrect dates — cross-reference with other records -- Different providers use different date formats — normalize to YYYY-MM-DD -- Nursing notes vs. physician notes may describe the same event differently — reconcile -- Lab results are time-stamped but dates on the page may differ — use the report date -- Missing entries are as important as present entries — document gaps in the chronology - -## Tags -practice_area: litigation -document_type: summary -skill_mode: summarization diff --git a/skills/legal/genesis-tort-case-summary/SKILL.md b/skills/legal/genesis-tort-case-summary/SKILL.md deleted file mode 100644 index 237b3f2e..00000000 --- a/skills/legal/genesis-tort-case-summary/SKILL.md +++ /dev/null @@ -1,38 +0,0 @@ -# Complex Tort Case Summary & Key Issue Identification - -## Prerequisites -- All available medical records and case documents -- Known injury type and claimed damages -- Defense theories or carrier denial rationales (if known) - -## Workflow -1. Analyze the full record set for tort-specific elements: - - Medical malpractice: standard of care, breach, causation, damages, expert requirements - - Products liability: defect type (design/manufacturing/warning), foreseeability - - Mass tort: exposure history, injury mechanism, commonality analysis - - General personal injury: duty, breach, causation, comparative fault -2. Identify key legal issues: - - Primary liability theories and supporting evidence - - Affirmative defenses raised or anticipated - - Comparative/contributory fault allocation signals - - Damages categories with record support -3. Produce a case summary document: - - Executive overview in 1–2 paragraphs - - Timeline of material events with source citations - - Strengths and weaknesses analysis for each claim element - - Damages summary with documented vs. non-documented categories - - Recommended expert witnesses and discovery priorities -4. Generate issue-spotting checklist for attorney review -5. Output a board-ready summary with all factual assertions traceable to records - -## Pitfalls -- Different tort types have different statute of limitations — check filing deadlines -- Expert witness requirements vary by jurisdiction — some states require pre-trial certification -- Document retention obligations begin at case evaluation — preserve everything -- Comparative fault rules differ by state — 50% bar, 51% bar, or pure comparative -- Don't conflate medical causation with legal causation — they require different proof - -## Tags -practice_area: litigation -document_type: summary -skill_mode: analysis diff --git a/skills/legal/habeas-case-strategy/SKILL.md b/skills/legal/habeas-case-strategy/SKILL.md deleted file mode 100644 index 5c7fc999..00000000 --- a/skills/legal/habeas-case-strategy/SKILL.md +++ /dev/null @@ -1,49 +0,0 @@ ---- -name: case-strategy-synthesis -language: en -description: >- - Synthesizes case holdings, relevant reasoning, and prevailing authorities - into a comprehensive case strategy document. Use when building litigation - strategy, identifying prevailing legal arguments, or creating a real-time - strategy memo from case law research. Trigger keywords: case strategy, - litigation strategy, prevailing argument, case law memo, legal reasoning - synthesis, strategy memo, case theory builder. -tags: - - analysis - - litigation - - research ---- - -# Case Strategy Synthesis - -Builds a litigation-ready strategy document from case law holdings, -reasoning patterns, and authoritative guidance. - -## Prerequisites -- Defined legal issue and jurisdiction -- Preliminary case facts (plaintiff/defendant posture, key claims) -- Research sources already retrieved (cases, statutes, regulations) - -## Workflow -1. **Issue Identification** — Distill the core legal questions from - the case facts and pleadings. -2. **Authority Mapping** — Chart governing cases, their holdings, and - the reasoning courts used to reach each holding. -3. **Precedent Hierarchy** — Rank authorities by binding strength - (supreme court > appellate > district), recency, and relevance. -4. **Argument Construction** — Build the strongest argument path using - the best authorities, anticipating counterarguments and distinguishing - adverse precedents. -5. **Weakness Assessment** — Identify gaps in the case law, unfavorable - precedents, and jurisdictional ambiguities. -6. **Strategy Memo Output** — Produce a structured document with - issues, authorities, arguments, counterarguments, and recommended - next steps (motions, expert retention, discovery needs). - -## Pitfalls -- Do not invent legal reasoning not supported by cited cases. -- Always distinguish between majority, concurring, and dissenting - opinions. -- Flag if the jurisdiction lacks directly on-point precedent. -- Avoid over-reliance on persuasive authority when binding authority - exists. diff --git a/skills/legal/habeas-citation-synthesis/SKILL.md b/skills/legal/habeas-citation-synthesis/SKILL.md deleted file mode 100644 index d35045ec..00000000 --- a/skills/legal/habeas-citation-synthesis/SKILL.md +++ /dev/null @@ -1,47 +0,0 @@ ---- -name: habeas-style-citation-synthesis -language: en -description: >- - Produces paragraph-level cited legal analysis grounded in primary - sources (case law, legislation, regulations). Use when the user needs - research with pinpoint citations, extracted passages, and source-level - verification — similar to Habeas.ai's approach. Trigger keywords: - cited legal analysis, paragraph citation, pinpoint citation, primary - source research, legislation citation, case law synthesis with quotes. -tags: - - research - - analysis - - litigation ---- - -# Cited Legal Synthesis (Habeas-Style) - -Generates legally-grounded analysis with paragraph-level citations to -primary sources, enabling source verification and reducing hallucination risk. - -## Prerequisites -- Defined legal question or issue -- Jurisdiction scope (state, federal, specific country) -- Preferred primary sources (case law, statutes, regulations) - -## Workflow -1. **Question Framing** — Convert the user's legal question into a - searchable query spanning cases, statutes, and regulations. -2. **Primary Source Retrieval** — Pull relevant authorities with - full-text excerpts, citing specific sections and paragraphs. -3. **Holding Extraction** — For each case, extract the specific - holding, reasoning, and any dicta that bears on the question. -4. **Citation-Linked Synthesis** — Compose the analysis with inline - citations to paragraph-level passages, allowing the reader to - verify every claim against the source. -5. **Conflicting Authority Check** — Identify and surface any - contradictory cases or overruled precedents. -6. **Verify Extracts** — Confirm that quoted text matches the source - document exactly. Flag any passages where the AI cannot verify. - -## Pitfalls -- Never fabricate citations or page/paragraph numbers. -- Do not paraphrase holdings without quoting the source language. -- If a question requires jurisdiction-specific rules not in scope, - state the limitation clearly. -- Distinguish between binding and persuasive authority explicitly. diff --git a/skills/legal/habeas-custom-agent/SKILL.md b/skills/legal/habeas-custom-agent/SKILL.md deleted file mode 100644 index 610a71b8..00000000 --- a/skills/legal/habeas-custom-agent/SKILL.md +++ /dev/null @@ -1,48 +0,0 @@ ---- -name: custom-legal-agent-builder -language: en -description: >- - Designs and deploys custom AI agents tailored to a specific practice - area and house style. Use when building practice-area-specific legal - assistants that produce outputs matching a firm's tone, formatting - standards, and analytical approach. Trigger keywords: custom legal AI, - practice area agent, house style AI, legal AI customization, firm-specific - assistant, branded legal research agent. -tags: - - research - - drafting - - legal-ops ---- - -# Custom Legal Agent Builder - -Architects practice-area-specific AI agents with firm-level tone, style, -and analytical preferences baked into the workflow. - -## Prerequisites -- Practice area definition (e.g., employment litigation, IP, family law) -- House style guide (tone, formatting, citation standards) -- Sample output documents (past memos, briefs, research reports) -- Firm-specific research databases or preferred sources - -## Workflow -1. **Practice Area Definition** — Map the scope of queries, document - types, and research sources the agent will handle. -2. **Style Alignment** — Ingest house style documents and past outputs - to calibrate tone, structure, citation format, and language preferences. -3. **Source Configuration** — Configure the agent to prioritize - firm-preferred sources (cases, statutes, internal precedents). -4. **Guardrails & Constraints** — Set limits on citation fabrication, - hallucination mitigation rules, and jurisdiction scope. -5. **Testing & Calibration** — Run the agent against sample questions, - compare output to partner-level work, adjust prompts and constraints. -6. **Deployment** — Integrate into the firm's workflow tools with - user-facing interface, access controls, and usage monitoring. - -## Pitfalls -- Do not over-constrain the agent — balance precision with flexibility - for novel legal questions. -- House style should cover citation format, tone, and structure but - not override legal accuracy. -- Regularly audit agent outputs for drift from style and accuracy. -- Always maintain attorney review gate for agent-generated content. diff --git a/skills/legal/haloo-brand-monitoring/SKILL.md b/skills/legal/haloo-brand-monitoring/SKILL.md deleted file mode 100644 index f67d8446..00000000 --- a/skills/legal/haloo-brand-monitoring/SKILL.md +++ /dev/null @@ -1,24 +0,0 @@ ---- -name: haloo-brand-monitoring -description: Design brand monitoring and enforcement strategies for trademark infringement across digital channels, marketplaces, and domain registrations. -tags: [intellectual-property, litigation] -skill_mode: [research, analysis] ---- - -# Brand Monitoring & Enforcement Strategy - -## Prerequisites -- Registered trademark portfolio with class definitions -- Knowledge of high-risk channels for the product category - -## Workflow -1. Map the trademark portfolio: registrations, pending applications, and known common-law uses -2. Identify monitoring targets: domain registrations, social media handles, marketplace listings, app stores -3. Define infringement indicators: identical marks, confusingly similar variants, cybersquatting patterns -4. Establish monitoring cadence and automated alert rules for new potentially infringing uses -5. Draft enforcement decision matrix: observation → cease & desist → opposition → litigation - -## Pitfalls -- Don't send cease-and-desist letters without confirming actual use in commerce -- Monitor both exact matches and phonetic/visual equivalents across all relevant classes -- Track enforcement outcomes to refine future monitoring thresholds diff --git a/skills/legal/haloo-portfolio-management/SKILL.md b/skills/legal/haloo-portfolio-management/SKILL.md deleted file mode 100644 index 49f8889d..00000000 --- a/skills/legal/haloo-portfolio-management/SKILL.md +++ /dev/null @@ -1,24 +0,0 @@ ---- -name: haloo-portfolio-management -description: Create and maintain trademark portfolio inventories with renewal tracking, usage documentation, and assignment records for global registrations. -tags: [intellectual-property, corporate] -skill_mode: [drafting, analysis] ---- - -# Trademark Portfolio Management - -## Prerequisites -- Complete trademark registration records across jurisdictions -- Internal records of trademark usage and licensing - -## Workflow -1. Inventory all registered marks with jurisdiction, filing date, registration date, and class assignments -2. Map usage evidence: product labels, marketing materials, advertising, website screenshots -3. Track renewal deadlines per jurisdiction with automated reminders (typically 10-year cycles) -4. Document all assignments, licenses, and security interests affecting each mark -5. Generate periodic portfolio health reports: upcoming deadlines, lapsed registrations, gap analysis - -## Pitfalls -- Never assume a registration is live without verifying current status — many lapse silently -- Document usage in the exact goods/services classes registered; over-broad use can weaken enforcement -- Track foreign filing deadlines (6-month priority) to avoid losing rights in new markets diff --git a/skills/legal/haloo-trademark-clearance/SKILL.md b/skills/legal/haloo-trademark-clearance/SKILL.md deleted file mode 100644 index fa16ad9b..00000000 --- a/skills/legal/haloo-trademark-clearance/SKILL.md +++ /dev/null @@ -1,24 +0,0 @@ ---- -name: haloo-trademark-clearance -description: Draft trademark clearance search strategies and results analysis covering USPTO, state registries, and common-law sources for new brand names. -tags: [intellectual-property, regulatory] -skill_mode: [research, analysis] ---- - -# Trademark Clearance Search Strategy - -## Prerequisites -- Proposed mark (word, design, or combination) -- Target goods/services classes and jurisdictions - -## Workflow -1. Generate variant spellings, phonetic equivalents, and common abbreviations of the proposed mark -2. Search USPTO TESS (federal), state registry databases, and WIPO Madrid for exact and similar marks -3. Query common-law sources: domain registries, social media, business directories, e-commerce platforms -4. Score similarity by mark appearance, sound, commercial impression, and goods/services overlap -5. Draft a clearance opinion summarizing findings, conflict risks, and recommended next steps - -## Pitfalls -- Don't ignore common-law trademarks — unregistered marks can still enforce rights -- Consider international classes beyond your immediate target; expansion risk matters -- Document every search source and date for traceability in opposition proceedings diff --git a/skills/legal/ip-portfolio-monitoring/SKILL.md b/skills/legal/ip-portfolio-monitoring/SKILL.md deleted file mode 100644 index b29d747b..00000000 --- a/skills/legal/ip-portfolio-monitoring/SKILL.md +++ /dev/null @@ -1,39 +0,0 @@ -name: ip-portfolio-monitoring -language: en -description: Monitors an IP portfolio across marketplaces and the open web for infringements — trademark abuse, counterfeit listings, unauthorized use of copyrighted images, and design patent violations. Generates prioritized enforcement reports with evidence packs ready for action. -tags: - - research - - analysis - - regulatory ---- - -# IP Portfolio Monitoring & Infringement Detection - -Systematically monitors a brand's intellectual property across ecommerce marketplaces, social platforms, and the open web for unauthorized use, counterfeiting, and infringement. - -## Prerequisites - -Before executing, collect: - -1. **IP register** — all trademarks, registered copyrights, design patents with registration numbers -2. **Authorized product images** — high-res product photos linked to SKUs/catalog items -3. **Authorized sellers list** — which distributors/resellers have permission -4. **Target marketplaces** — Amazon, Shopify, eBay, Etsy, TikTok Shop, Temu, Shein, etc. - -## Workflow - -1. **Catalog ingestion** — import product images, descriptions, and IP register -2. **Computer vision scan** — deploy visual search against target marketplaces using product imagery -3. **Trademark keyword search** — scan listings for unauthorized use of registered marks -4. **Result correlation** — match findings against authorized sellers to flag unauthorized use -5. **Evidence packaging** — compile screenshot evidence, URLs, timestamps, and IP matching results -6. **Prioritization** — rank by severity (direct counterfeiting > unauthorized reselling > image theft) -7. **Enforcement handoff** — output structured report ready for takedown or legal action - -## Pitfalls - -- Computer vision false positives from similar but legitimate products -- Listing text evasion (blurred images, altered trademarks, keyword stuffing) -- Platform detection gaps — some marketplaces lack robust enforcement APIs -- Local marketplace variants (e.g., regional .cn domains) may escape detection -- Results noise requires human review before enforcement action diff --git a/skills/legal/lawxy-case-sentiment-analysis/SKILL.md b/skills/legal/lawxy-case-sentiment-analysis/SKILL.md deleted file mode 100644 index dd9ad491..00000000 --- a/skills/legal/lawxy-case-sentiment-analysis/SKILL.md +++ /dev/null @@ -1,34 +0,0 @@ ---- -name: case-sentiment-analysis -description: Analyze case files to determine judicial and opposing-party sentiment, predict case trajectory, and identify key issues from pleadings, briefs, and evidence -tags: - - practice_area: litigation - - document_type: brief, pleading - - skill_mode: analysis ---- - -# Legal Case Sentiment Analysis - -Analyze litigation files to extract sentiment signals, predict likely outcomes, and surface critical strategic issues. - -## Prerequisites - -- Full case file (pleadings, briefs, motions, deposition excerpts) -- Jurisdiction and court context -- Opposing party profile (if available) - -## Workflow - -1. Ingest all case documents and build a complete matter timeline -2. Score sentiment per party, per judge, and per issue (strongly-positive → strongly-negative) -3. Identify pivotal moments: key motions filed, adverse rulings, evidentiary wins/losses -4. Map sentiment shifts across the litigation lifecycle -5. Produce a case narrative with predictive outcome ranges (settlement likelihood, trial probability, estimated damages) -6. Recommend strategic adjustments based on sentiment trajectory - -## Pitfalls - -- Sentiment ≠ legal merit; a positive sentiment score doesn't guarantee a win -- Judge-specific bias must account for published rulings history, not just current case -- Avoid overfitting to a small sample of the judge's past decisions -- Flag when sentiment analysis is unreliable (unclear authorship, ambiguous language) diff --git a/skills/legal/lawxy-contract-review-studio/SKILL.md b/skills/legal/lawxy-contract-review-studio/SKILL.md deleted file mode 100644 index 6084a495..00000000 --- a/skills/legal/lawxy-contract-review-studio/SKILL.md +++ /dev/null @@ -1,34 +0,0 @@ ---- -name: contract-review-studio -description: AI-powered contract review in Word with fallback language suggestions, redlining, and contextual Q&A on revisions -tags: - - practice_area: transactional - - document_type: agreement - - skill_mode: analysis ---- - -# Contract Review Studio — AI Redlining & Fallback Language - -AI-assisted contract review directly in Microsoft Word with intelligent redline suggestions, fallback clauses, and contextual revision explanations. - -## Prerequisites - -- Contract in Word format (.docx) -- Client's standard clause library or redline preferences -- Counterparty context (who they are, what they typically push for) - -## Workflow - -1. Open contract in the Word add-in — AI scans the document for risky, non-standard, or missing clauses -2. Flag each issue with a confidence score and suggested revision -3. Provide fallback language for flagged clauses (standard alternatives the user can accept with one click) -4. Generate a redline comparing current contract against a baseline template -5. Support contextual Q&A: ask questions about specific clauses and get cited answers -6. Export the reviewed document with all changes tracked - -## Pitfalls - -- Fallback language must be jurisdiction-appropriate — never use generic boilerplate for specialized clauses -- Confidence scores are guidance, not decisions — always review high-risk flags manually -- Redlines should explain *why* a clause is flagged, not just *that* it is flagged -- The Word add-in may not preserve complex formatting — verify output before sending diff --git a/skills/legal/lawxy-due-diligence/SKILL.md b/skills/legal/lawxy-due-diligence/SKILL.md deleted file mode 100644 index ff39fc14..00000000 --- a/skills/legal/lawxy-due-diligence/SKILL.md +++ /dev/null @@ -1,34 +0,0 @@ ---- -name: multi-document-due-diligence -description: Process hundreds of documents to extract obligations, identify risks, and produce a structured due diligence report with flagged items -tags: - - practice_area: corporate - - document_type: analysis, summary, checklist - - skill_mode: analysis ---- - -# Multi-Document Due Diligence Intelligence - -Analyze large document sets (contracts, financials, compliance docs) to surface material obligations, risks, and anomalies. - -## Prerequisites - -- Complete document set (contracts, exhibits, schedules, corporate records) -- Deal context (transaction type, parties, materiality thresholds) -- Redline comparison documents if available - -## Workflow - -1. Ingest all documents and build a unified knowledge graph of parties, obligations, and relationships -2. Extract key terms: payment schedules, termination clauses, change-of-control, indemnities, covenants -3. Score each obligation by risk level (high/medium/low) based on materiality thresholds and deal context -4. Identify anomalies: missing signatures, inconsistent dates, conflicting clauses across documents -5. Generate a structured diligence report with all flagged items mapped to source documents -6. Provide recommended actions: renegotiate, waive, disclose, or accept - -## Pitfalls - -- Materiality thresholds must be deal-specific; generic thresholds miss context -- Never assume a clause is standard without comparing to market norms -- Conflicting obligations across multiple agreements must be surfaced, not silently resolved -- Financial document analysis requires domain-specific number parsing (not generic NLP) diff --git a/skills/legal/lawxy-legal-agent-network/SKILL.md b/skills/legal/lawxy-legal-agent-network/SKILL.md deleted file mode 100644 index 8b7adffe..00000000 --- a/skills/legal/lawxy-legal-agent-network/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ ---- -name: legal-agent-network -description: Orchestrate specialized AI agents to draft, redline, and research legal documents under human control with rule-based governance -tags: - - practice_area: corporate - - document_type: agreement, letter - - skill_mode: drafting ---- - -# Legal Agent Network — Multi-Agent Drafting & Research - -Coordinate autonomous AI agents to handle parallel legal tasks under controlled rules, with human oversight at each gate. - -## Prerequisites - -- Clear task decomposition (what agents do, in what order) -- Human-in-the-loop approval gates defined -- Agent role definitions (researcher, drafter, redliner, compliance checker) -- Access to relevant precedent and firm knowledge base - -## Workflow - -1. Define the task graph: which agents run, which depend on which outputs -2. Launch parallel agents for independent subtasks (e.g., research + contract review simultaneously) -3. Each agent works under explicit rules: acceptable clauses, prohibited language, citation requirements -4. Agent outputs are collected and presented for human review at each gate -5. Human approves, rejects, or modifies outputs before advancing to next stage -6. Final assembled document incorporates all approved agent outputs with full audit trail - -## Pitfalls - -- Agents must never auto-advance past a human gate without explicit approval -- Rule conflicts between agents (e.g., drafter suggests a clause compliance checker flags) must surface to human -- Over-parallelization can create conflicting drafts; serialize when task dependencies exist -- Always maintain agent provenance in the final document for auditability diff --git a/skills/legal/lawy-contract-intelligence/SKILL.md b/skills/legal/lawy-contract-intelligence/SKILL.md deleted file mode 100644 index 9e1cb930..00000000 --- a/skills/legal/lawy-contract-intelligence/SKILL.md +++ /dev/null @@ -1,60 +0,0 @@ ---- -name: contract-intelligence -language: en -description: >- - Analyzes contracts for risk, non-standard terms, and negotiation leverage - points. Reviews agreements against client playbooks and identifies clauses - that deviate from acceptable standards. Use when reviewing incoming - contracts, preparing for negotiations, or conducting portfolio reviews. -tags: - - contracts - - analysis - - research ---- - -# Contract Intelligence - -Systematic analysis of contracts for risk identification, non-standard term -detection, and negotiation guidance. - -## Prerequisites - -- Contract text (complete, not truncated) -- Client playbooks or acceptable terms (if available) -- Counterparty identity and relationship context -- Deal context (value, duration, strategic importance) - -## Workflow - -1. **Structural analysis** — map the contract's architecture: - - Identify all clause types and their locations - - Note any missing standard sections (indemnification, limitation of liability) - - Flag unusual structures or non-standard organization -2. **Risk clause identification** — scan for high-risk provisions: - - Indemnification obligations and carve-outs - - Limitation of liability (caps, exclusions, carve-outs) - - Termination rights (cause, convenience, notice periods) - - IP ownership and license grants - - Confidentiality scope and duration - - Data protection and privacy provisions -3. **Playbook deviation check** — compare against client's acceptable terms: - - Flag terms that are beyond acceptable parameters - - Score deviation severity (critical/high/medium/low) - - Suggest acceptable alternative language -4. **Negotiation intelligence** — assess leverage: - - Market-standard position on key terms - - Counterparty's likely bargaining positions - - Tradeable concessions and their relative value -5. **Output** — deliver a contract review memo with: - - Executive summary of key risks - - Clause-by-clause analysis with risk scores - - Recommended negotiation strategy and fallback positions - - Redlined suggested language for critical terms - -## Pitfalls - -- Always read the entire contract — key risks are often in definitions or cross-referenced sections -- Context matters enormously: the same clause is low-risk in an NDA and high-risk in an MSA -- Market standards vary by industry and deal size — don't apply startup templates to enterprise deals -- The most dangerous clauses are ones that look standard but have subtle modifications -- Never rely solely on clause-level analysis — the interplay between provisions can create emergent risk diff --git a/skills/legal/lawy-matter-summarization/SKILL.md b/skills/legal/lawy-matter-summarization/SKILL.md deleted file mode 100644 index 835bab34..00000000 --- a/skills/legal/lawy-matter-summarization/SKILL.md +++ /dev/null @@ -1,58 +0,0 @@ ---- -name: matter-summarization -language: en -description: >- - Produces comprehensive matter summaries from case files, documents, and - communications. Converts voluminous matter data into instant context for - legal work including deadlines, key parties, procedural history, and - outstanding tasks. Use when onboarding to a new matter or preparing for - a meeting where full context is needed quickly. -tags: - - analysis - - summarization ---- - -# Matter Summarization - -Process for converting voluminous matter data into a structured, actionable -summary that establishes instant context for legal work. - -## Prerequisites - -- Matter file contents (pleadings, correspondence, documents, notes) -- Client engagement letter or matter opening memorandum -- Known deadlines and procedural calendar - -## Workflow - -1. **Document ingestion and classification** — categorize all matter documents: - - Pleadings (complaint, answer, motions, orders) - - Correspondence (client emails, opposing counsel letters) - - Court filings (hearings, appearances, deadlines) - - Evidence (witness statements, exhibits, expert reports) - - Administrative (invoices, timesheets, billing records) -2. **Key entity extraction** — identify: - - Parties (plaintiff, defendant, intervenors, witnesses) - - Counsel (lead, associates, co-counsel, opposing) - - Courts and judges - - Dates (filing, service, hearings, deadlines) -3. **Procedural timeline construction** — build a chronological docket: - - Key filings and their dates - - Court-ordered deadlines and response periods - - Upcoming hearings and trial dates - - Statute of limitations exposure -4. **Status assessment** — summarize the matter's current position: - - Procedural posture (pre-filing, discovery, motion practice, trial, appeal) - - Outstanding issues and pending motions - - Key risks and leverage points - - Recommended next steps -5. **Output delivery** — produce the matter summary in a structured format - suitable for quick comprehension and client communication - -## Pitfalls - -- Don't assume document metadata is accurate — verify filing dates against court records -- Calendar items from opposing counsel are often unreliable — cross-reference with court rules -- Matter summaries become stale quickly — establish a refresh cadence tied to docket events -- Privilege boundaries: ensure the summary doesn't inadvertently create new privilege issues -- Summarization loses nuance — always flag high-risk items for attorney review before client delivery diff --git a/skills/legal/lawy-verified-legal-research/SKILL.md b/skills/legal/lawy-verified-legal-research/SKILL.md deleted file mode 100644 index f7d28c08..00000000 --- a/skills/legal/lawy-verified-legal-research/SKILL.md +++ /dev/null @@ -1,54 +0,0 @@ ---- -name: verified-legal-research -language: en -description: >- - Conducts legal research with optional human lawyer verification. Combines - AI-generated research answers with a lawyer-in-the-loop verification step - for jurisdiction-specific accuracy. Use when research accuracy is critical - and verification can reduce hallucination risk. -tags: - - research - - drafting - - analysis ---- - -# Verified Legal Research - -Legal research workflow that couples AI-generated answers with human verification -for jurisdiction-specific accuracy and confidence. - -## Prerequisites - -- Clear research question (single issue, scoped to one jurisdiction) -- Identified governing law and authority sources -- Knowledge that verification is available for the target jurisdiction - -## Workflow - -1. **AI research generation** — produce an initial research answer: - - Identify the legal issue and applicable jurisdiction - - Retrieve primary authority (statutes, regulations, case law) - - Supplement with secondary authority and practice guidance - - Draft the answer with inline source references -2. **Self-audit** — review the AI-generated answer: - - Are all citations current and still good law? - - Does the answer address the precise question asked? - - Are there competing authorities not mentioned? - - Is the jurisdiction correctly scoped? -3. **Verification handoff** (optional) — for questions requiring verified accuracy: - - Submit the research question and AI-generated answer to a local practicing attorney - - Attorney reviews the answer against the source authorities - - Attorney provides corrections, additions, or confirmation -4. **Final delivery** — produce the verified research output: - - State whether verification was performed - - If verified, include the verifier's modifications and confidence level - - If not verified, include standard AI accuracy disclaimer - - Provide all source links and a research trail - -## Pitfalls - -- Never present unverified AI research as authoritative without clear disclaimers -- Verification doesn't guarantee correctness — verify the verifier's answer against sources when possible -- Jurisdiction is the most common error vector — confirm venue rules and local procedures -- AI can hallucinate citations that look real but don't exist — always validate every citation independently -- The lawyer-in-the-loop adds latency — for routine research, the AI answer with self-audit may be sufficient diff --git a/skills/legal/legal-translation/SKILL.md b/skills/legal/legal-translation/SKILL.md deleted file mode 100644 index 13285a09..00000000 --- a/skills/legal/legal-translation/SKILL.md +++ /dev/null @@ -1,29 +0,0 @@ ---- -name: Legal Document Translation -description: Translate legal documents between English and 12+ Indian vernacular languages with jurisdictional awareness. Maintains legal precision while preserving tone and formal register. Includes option for human review. -tags: - - practice_area: litigation - - document_type: pleading, brief, letter - - skill_mode: drafting ---- - -# Legal Document Translation - -## Prerequisites -- Source legal document (pleading, brief, submission, contract) -- Target language(s) from supported set (English + 12 Indian vernaculars) -- Jurisdiction context (affects formal register and terminology) - -## Workflow -1. **Identify source and target languages** — determine document type and supported translation pairs -2. **Parse legal terminology** — map jurisdiction-specific terms to target-language equivalents -3. **Translate structure** — preserve document format, citations, headings, and formal register -4. **Review legal precision** — ensure technical terms, citations, and procedural references are accurate -5. **Human review flag** — mark sections where human expert review is recommended for nuance -6. **Deliver bilingual output** — produce parallel or target-language-only document as needed - -## Pitfalls -- Legal terminology does not always have direct translations — use jurisdictionally appropriate terms -- Never auto-translate statutory references; preserve original citations -- Tone matters: formal legal register must be maintained, not casual language -- Recommend human review for documents with high stakes or complex procedural language diff --git a/skills/legal/legal1up-discovery/SKILL.md b/skills/legal/legal1up-discovery/SKILL.md deleted file mode 100644 index e2d9527c..00000000 --- a/skills/legal/legal1up-discovery/SKILL.md +++ /dev/null @@ -1,48 +0,0 @@ ---- -name: federal-discovery-analyst -language: en -description: >- - Analyzes voluminous discovery materials for federal litigation using - AI-driven document processing with human expert oversight. Use when - handling plaintiff-side federal cases needing discovery support, - exhibit preparation, or deposition question guides. Trigger keywords: - federal discovery, deposition exhibit, discovery analysis, discovery - request assistance, plaintiff-side discovery, deposition questions, - exhibit preparation, joint appendix. -tags: - - analysis - - litigation - - e-discovery ---- - -# Federal Discovery Analyst (AI + Human) - -Produces AI-driven discovery analysis with paralegal-level accuracy -and human expert review for federal litigation cases. - -## Prerequisites -- Discovery materials (interrogatories, RFPs, depositions, privilege - logs, expert reports) -- Case type and jurisdiction (federal court, specific district) -- Key witness/deponent list and timeline - -## Workflow -1. **Document Ingestion** — Process uploaded discovery materials, - classify by type, and index searchable content. -2. **Thematic Analysis** — Group documents by issue, identify - inconsistencies across testimonies and written responses. -3. **Fact Extraction** — Pull key facts, dates, admissions, and - contradictions into a structured fact matrix. -4. **Deposition Guide Creation** — Generate deposition questions - targeting inconsistencies, admissions, and credibility issues. -5. **Exhibit Preparation** — Compile exhibits for trial or motions - with Bates numbering, cross-references, and witness attribution. -6. **Human Review Gate** — All outputs routed to expert legal - professionals for accuracy verification before attorney use. - -## Pitfalls -- AI analysis is a first pass — human review is mandatory for all - discovery outputs in federal court. -- Do not present AI-inferred facts as established evidence. -- Preserve chain of custody for all documents. -- Flag privilege concerns when analyzing mixed document sets. diff --git a/skills/legal/legal1up-joint-appendix/SKILL.md b/skills/legal/legal1up-joint-appendix/SKILL.md deleted file mode 100644 index b33a8c51..00000000 --- a/skills/legal/legal1up-joint-appendix/SKILL.md +++ /dev/null @@ -1,48 +0,0 @@ ---- -name: joint-appendix-compiler -language: en -description: >- - Compiles and formats joint appendices for federal appellate court - filing. Use when preparing the official record on appeal, organizing - district court filings, orders, and exhibits into a court-compliant - joint appendix. Trigger keywords: joint appendix, appellate record, - appendix compilation, federal appellate, record on appeal, FRAP 30, - appellate brief appendix. -tags: - - drafting - - litigation - - appellate ---- - -# Joint Appendix Compiler - -Produces court-compliant joint appendices for federal appellate proceedings. - -## Prerequisites -- District court docket and filed documents -- Appellate court rules (circuit-specific formatting) -- Parties' stipulation on what to include in the appendix -- FRAP 30 requirements and circuit supplement rules - -## Workflow -1. **Docket Review** — Pull the complete district court docket and - identify all documents required by FRAP 30(a)(1) and circuit rules. -2. **Content Selection** — Work with counsel to select opinions, - orders, rulings, and exhibits that the appealing party deems - necessary for the appellate court's review. -3. **Ordering & Pagination** — Arrange documents in the required - sequence: opinion/order first, then appendices, with correct - pagination (appendix pages numbered separately). -4. **Format Compliance** — Ensure font, margin, and binding - requirements match the specific circuit's rules. -5. **Cross-Reference Index** — Create a table of contents with - appendix page numbers for easy citation during briefing. -6. **Quality Check** — Verify all documents are correctly captured, - paginated, and bound per local rules. - -## Pitfalls -- Missing a required document can result in dismissal or remand. -- Appendix pages must be numbered separately from brief pages. -- Check circuit-specific supplement rules (e.g., 9th Cir. Rule 30-1). -- Do not include documents not relied upon in the briefs. -- Verify all exhibits referenced in the brief are in the appendix. diff --git a/skills/legal/litigation-drafting/SKILL.md b/skills/legal/litigation-drafting/SKILL.md deleted file mode 100644 index 50724360..00000000 --- a/skills/legal/litigation-drafting/SKILL.md +++ /dev/null @@ -1,29 +0,0 @@ ---- -name: Litigation Document Drafting -description: Generate polished first drafts of litigation documents — motions, briefs, responses, discovery requests — guided by jurisdiction, facts, and prior filings. Mirrors attorney voice and follows procedural rules. -tags: - - practice_area: litigation - - document_type: pleading, motion, brief, letter - - skill_mode: drafting ---- - -# Litigation Document Drafting - -## Prerequisites -- Case facts and relevant prior filings (pleadings, motions, discovery) -- Jurisdiction and court rules -- Attorney writing style preferences (optional) - -## Workflow -1. **Gather inputs** — collect case facts, jurisdiction, prior filings, and writing style if available -2. **Select document type** — motion, brief, response, discovery request, demand letter -3. **Draft structure** — generate section headings aligned with jurisdictional requirements and court rules -4. **Fill with arguments** — incorporate case-specific facts, legal theories, and citations -5. **Style match** — apply attorney's voice, tone, and formatting conventions -6. **Export** — produce Word-compatible output with formatting intact - -## Pitfalls -- Do NOT invent case facts or citations — only use provided materials -- Jurisdictional format requirements vary; always verify against local rules -- Never use open-model data; this is litigation work requiring accuracy -- Ensure no client data is used to train external models diff --git a/skills/legal/med-mal-case-evaluation/SKILL.md b/skills/legal/med-mal-case-evaluation/SKILL.md deleted file mode 100644 index 12dd6bd5..00000000 --- a/skills/legal/med-mal-case-evaluation/SKILL.md +++ /dev/null @@ -1,30 +0,0 @@ ---- -name: Medical Malpractice Case Evaluation -description: Clinically rigorous analysis of medical malpractice cases — evaluates standard of care, breach, causation, and damages using clinical expertise. Reduces case evaluation time from days to minutes. -tags: - - practice_area: litigation - - document_type: analysis, summary, memo - - skill_mode: analysis ---- - -# Medical Malpractice Case Evaluation - -## Prerequisites -- Medical records (chart notes, imaging reports, lab results) -- Incident timeline and patient history -- Plaintiff or defense perspective -- Relevant jurisdiction's standard of care - -## Workflow -1. **Review medical records** — extract clinical events, treatments, and outcomes chronologically -2. **Identify standard of care** — determine what a reasonably prudent provider would have done in the same circumstances -3. **Assess breach** — evaluate whether the provider's actions fell below the standard -4. **Analyze causation** — determine whether breach directly caused or contributed to harm -5. **Evaluate damages** — quantify medical costs, lost wages, pain and suffering -6. **Assign merit score** — produce data-driven case merit assessment for early decision-making - -## Pitfalls -- Clinical accuracy is paramount — this is not generic legal analysis; medical facts drive everything -- Do not confuse correlation with causation in medical outcomes -- Standard of care varies by specialty, geography, and time period -- Always flag where expert witness testimony is required diff --git a/skills/legal/ml-contract-provision-extraction/SKILL.md b/skills/legal/ml-contract-provision-extraction/SKILL.md deleted file mode 100644 index 83e45afe..00000000 --- a/skills/legal/ml-contract-provision-extraction/SKILL.md +++ /dev/null @@ -1,38 +0,0 @@ -name: ml-contract-provision-extraction -language: en -description: Uses machine learning to extract key provisions from contracts at scale — payment terms, termination clauses, indemnity obligations, exclusivity, liability caps, and change-of-control. Produces structured data suitable for downstream analysis, reporting, or risk scoring. -tags: - - analysis - - agreement - - corporate ---- - -# ML-Powered Contract Provision Extraction - -Automates extraction of material contract provisions using machine learning models trained on legal document patterns, producing structured data for downstream analytics. - -## Prerequisites - -Before executing, collect: - -1. **Contract document(s)** — executed or draft agreements in PDF, DOCX, or text format -2. **Provision categories** — which provisions to extract (payment terms, termination, indemnity, liability, exclusivity, etc.) -3. **Output format** — structured JSON, spreadsheet columns, or report format -4. **Review scope** — single contract, portfolio batch, or continuous ingestion pipeline - -## Workflow - -1. **Document ingestion** — parse contract text, handle multi-page formatting, OCR if needed -2. **Clause segmentation** — identify and separate individual clauses from the contract body -3. **Provision classification** — ML model classifies each clause into provision categories -4. **Key data extraction** — extract specific values: dates, amounts, thresholds, named parties -5. **Structure output** — produce structured representation with clause text, metadata, and confidence scores -6. **Quality review** — flag low-confidence extractions for human review -7. **Integration** — export to contract management system, analytics dashboard, or data warehouse - -## Pitfalls - -- ML confidence varies by contract quality — poor formatting or non-standard language reduces accuracy -- Cross-references to other clauses need resolution — ML may extract clause text without contextual links -- Jurisdiction-specific provisions (UCC vs common law) may not be in training set -- Extracted data should always be validated before use in negotiations or compliance reporting diff --git a/skills/legal/pactum-negotiation-agent/SKILL.md b/skills/legal/pactum-negotiation-agent/SKILL.md deleted file mode 100644 index 170b8494..00000000 --- a/skills/legal/pactum-negotiation-agent/SKILL.md +++ /dev/null @@ -1,50 +0,0 @@ ---- -name: procurement-negotiation-agent -language: en -description: >- - Orchestrates AI procurement negotiation agents across supplier categories, - price lists, discounts, payment terms, and rebates. Use when designing - or implementing autonomous procurement negotiation workflows with - policy guardrails. Trigger keywords: procurement negotiation AI, supplier - negotiation agent, purchasing agent, procurement AI, contract negotiation - agents, price negotiation, procurement automation, tactical sourcing. -tags: - - contractual - - analysis - - corporate ---- - -# Procurement Negotiation Agent - -Architects autonomous AI negotiation agents for procurement with policy -guardrails and human-in-the-loop approval gates. - -## Prerequisites -- Procurement policy document (spending limits, approved vendors, - pricing benchmarks, discount tiers) -- Supplier catalog or purchase order history -- Category definitions and spend analytics - -## Workflow -1. **Policy Configuration** — Load procurement guardrails: approval - thresholds, preferred pricing, payment terms, and discount policies. -2. **Supplier Mapping** — Index all suppliers, categories, and historical - contract terms into the agent's knowledge base. -3. **Agent Assignment** — Route negotiations to the appropriate agent - type (requisition alignment, tactical sourcing, price list, discount, - payment terms, rebate). -4. **Negotiation Execution** — Agent engages suppliers autonomously - within guardrails, or routes to human approval when exceeding limits. -5. **Outcome Tracking** — Record negotiated terms, savings achieved, - supplier satisfaction scores, and compliance flags. -6. **Continuous Learning** — Update pricing benchmarks and negotiation - playbooks based on realized outcomes. - -## Pitfalls -- Never allow an agent to exceed configured approval thresholds - without human review. -- Supplier relationship risk must be monitored — aggressive - negotiation can damage long-term partnerships. -- Track all negotiated terms with version control for audit trail. -- Ensure data privacy: supplier pricing data must not be shared - across competing organizations. diff --git a/skills/legal/pactum-price-list/SKILL.md b/skills/legal/pactum-price-list/SKILL.md deleted file mode 100644 index 560fb11a..00000000 --- a/skills/legal/pactum-price-list/SKILL.md +++ /dev/null @@ -1,48 +0,0 @@ ---- -name: price-list-negotiation-policy -language: en -description: >- - Designs and maintains continuous price-list negotiation policies for - procurement teams. Use when establishing automated price enforcement, - index-linked pricing, or market-condition pricing updates across - supplier contracts. Trigger keywords: price list negotiation, continuous - pricing, index-linked pricing, market pricing, procurement pricing - policy, automated price updates, supplier pricing enforcement. -tags: - - regulatory - - analysis - - contractual ---- - -# Price List Negotiation Policy - -Establishes automated pricing enforcement policies that update contract -terms continuously based on market conditions and indices. - -## Prerequisites -- Current supplier contracts with pricing clauses -- Market indices or benchmarks relevant to the category -- Price update triggers and thresholds -- Approval workflow for pricing changes - -## Workflow -1. **Index Mapping** — Identify which market indices or benchmarks - apply to each supplier category (commodities, labor, freight, etc.). -2. **Formula Definition** — Define price adjustment formulas linking - contract prices to indices (e.g., CPI, PPI, commodity futures). -3. **Trigger Configuration** — Set thresholds for when price changes - are auto-applied vs. requiring supplier notification or approval. -4. **Monitoring Engine** — Continuously track index values and compute - adjusted prices against current contract terms. -5. **Change Execution** — Auto-generate price update notifications to - suppliers, or create revised POs within approved parameters. -6. **Audit & Reporting** — Log all pricing changes, supplier responses, - and realized savings for finance and compliance review. - -## Pitfalls -- Price adjustments must comply with existing contract terms — do - not override force majeure or price freeze clauses. -- Sudden index spikes require manual review before auto-execution. -- Supplier contracts may contain most-favored-nation clauses that - limit pricing flexibility. -- Always maintain a complete audit trail for financial reporting. diff --git a/skills/legal/patent-infringement-claim-chart/SKILL.md b/skills/legal/patent-infringement-claim-chart/SKILL.md deleted file mode 100644 index 53cf0554..00000000 --- a/skills/legal/patent-infringement-claim-chart/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ ---- -name: patent-infringement-claim-chart -description: Generate AI-assisted patent claim charts mapping patent claims to product features -tags: - practice: litigation - document: analysis - mode: drafting ---- - -# Patent Infringement Claim Chart - -Generate structured claim charts comparing patent claim elements against target product features. - -## Prerequisites - -- Target patent number and claims (especially Claim 1) -- Target product description, specifications, or source code -- Product testing results or teardown documentation - -## Workflow - -1. **Extract claim elements** — Break down each claim into individual elements (element-by-element decomposition) -2. **Map to product features** — For each claim element, identify the corresponding product feature or component -3. **Gather evidence** — Cite specific product documentation, screenshots, test results, or source code lines -4. **Score confidence** — Assign a confidence score (0–100) for each element mapping based on evidence quality -5. **Generate analysis** — Write AI-powered analysis explaining why each element is met or not met -6. **Summary scoring** — Aggregate element scores into an overall infringement assessment - -## Pitfalls - -- Don't skip dependency chains — dependent claims add further limitations -- Element decomposition must be precise — vague elements lead to weak charts -- Product evidence must be specific (cite exact version, date, section) -- Missing elements can defeat infringement — flag gaps prominently -- Design patent claim charts use drawings, not text claims — adapt approach accordingly diff --git a/skills/legal/patent-infringement-detection/SKILL.md b/skills/legal/patent-infringement-detection/SKILL.md deleted file mode 100644 index 3c344e1a..00000000 --- a/skills/legal/patent-infringement-detection/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ ---- -name: patent-infringement-detection -description: Identify potential patent infringements by screening products against patent portfolios -tags: - practice: litigation - document: analysis - mode: analysis ---- - -# Patent Infringement Detection - -Automated screening of products and services against patent portfolios to identify infringement targets. - -## Prerequisites - -- Patent portfolio (own or client's) with key claims identified -- Product catalog or service descriptions to screen -- Target industry or competitor list - -## Workflow - -1. **Product surveillance** — Continuously scan product releases, software updates, and service changes -2. **Claim-element matching** — Map each product feature against relevant patent claims -3. **Competitor tracking** — Monitor competitor products for infringement patterns -4. **Portfolio analysis** — Rank patents by infringement likelihood and commercial value -5. **Risk scoring** — Generate quantitative IP risk scores for each product-patent pair -6. **Prioritization report** — Output ranked list of highest-value enforcement targets - -## Pituts - -- Software patents require careful claim construction — software can implement same function differently -- Don't overlook indirect infringement (inducement, contributory) alongside direct infringement -- Prior art analysis should precede enforcement to assess patent strength -- International markets may have separate patent coverage — verify territorial scope -- Open-source components may introduce third-party IP complications diff --git a/skills/legal/patent-invention-disclosure/SKILL.md b/skills/legal/patent-invention-disclosure/SKILL.md deleted file mode 100644 index 2cdd302f..00000000 --- a/skills/legal/patent-invention-disclosure/SKILL.md +++ /dev/null @@ -1,37 +0,0 @@ ---- -name: patent-invention-disclosure -description: Draft structured invention disclosure documents for patent prosecution -tags: - - practice: patent - - document: drafting - - mode: drafting ---- - -# Patent Invention Disclosure - -Create comprehensive invention disclosure documents that enable strong patent claims. - -## Prerequisites -- Inventor interviews or meeting notes -- Technical drawings or schematics -- Understanding of the problem being solved -- Knowledge of closest known prior art - -## Workflow -1. **Structure**: Organize disclosure into standard sections: - - Technical field and background - - Problem statement and prior art deficiencies - - Summary of the invention - - Detailed description with reference numerals - - Brief description of drawings - - Claims (provisional) - - Advantages over prior art -2. **Enrich**: Fill gaps by searching for related patents and publications -3. **Claims**: Draft independent and dependent claims covering key embodiments -4. **Review**: Verify enablement — would a POSITA practice the invention from this disclosure? - -## Pitfalls -- Don't use marketing language; technical specificity wins -- Include alternative embodiments — they become fallback claims -- Number all figures and reference consistently in the description -- Explicitly state what the prior art does NOT do diff --git a/skills/legal/patent-portfolio-analysis/SKILL.md b/skills/legal/patent-portfolio-analysis/SKILL.md deleted file mode 100644 index 67bd7d9d..00000000 --- a/skills/legal/patent-portfolio-analysis/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ ---- -name: patent-portfolio-analysis -description: Analyze patent portfolio value, overlap, and enforcement strategy recommendations -tags: - practice: corporate - document: analysis - mode: analysis ---- - -# Patent Portfolio Analysis - -Strategic analysis of patent portfolios for valuation, monetization, and enforcement planning. - -## Prerequisites - -- Patent portfolio listing (patent numbers, titles, issue dates, statuses) -- Business context (core products, target markets, competitive landscape) -- Monetization goals (licensing, enforcement, defense, sale) - -## Workflow - -1. **Portfolio inventory** — Catalog all patents with key metadata (claims, status, maintenance fees, expiration) -2. **Technology clustering** — Group patents by technology area and product relevance -3. **Strength assessment** — Evaluate each patent's enforceability, claim breadth, and prior art exposure -4. **Commercial mapping** — Map patents to products, revenue streams, and market segments -5. **Monetization scoring** — Rank patents by licensing potential, enforcement likelihood, and defense value -6. **Strategic recommendations** — Prioritize maintenance, filing continuation applications, or settlement targets - -## Pitfalls - -- Maintenance fee deadlines are absolute — missing them kills the patent -- Continuation applications may be time-sensitive — check terminal disclaimers -- Patent term adjustment (PTA) and extension (PTE) can significantly alter expiration dates -- Overlapping patents within a portfolio can be redundant — identify and consolidate -- Defensive publications may undermine patent value if not properly tracked diff --git a/skills/legal/patentability-assessment/SKILL.md b/skills/legal/patentability-assessment/SKILL.md deleted file mode 100644 index bbdb7617..00000000 --- a/skills/legal/patentability-assessment/SKILL.md +++ /dev/null @@ -1,32 +0,0 @@ ---- -name: patentability-assessment -description: Preliminary patentability assessment evaluating novelty and non-obviousness before formal prosecution -tags: - - practice: patent - - document: analysis - - mode: analysis ---- - -# Patentability Assessment - -Provide a preliminary assessment of whether an invention likely meets patentability requirements. - -## Prerequisites -- Invention disclosure document (PDF, DOC, DOCX) -- Key technical features description -- Competing products or prior art known to the inventor - -## Workflow -1. **Parse**: Extract technical features from invention disclosure -2. **Search**: Conduct broad patentability search across patent and NPL databases -3. **Evaluate**: For each feature, assess: - - Is it novel? (no single reference discloses all elements) - - Is it non-obvious? (would a POSITA find the combination obvious) -4. **Score**: Assign patentability score (high/medium/low) with rationale -5. **Recommend**: Suggest claim strategies to overcome identified references - -## Pitfalls -- This is preliminary, not a formal legal opinion — mark it as such -- Don't claim definitively that something IS or IS NOT patentable -- Focus on the combination of features, not individual elements -- Be honest about close prior art — it's better to flag it early diff --git a/skills/legal/patentprior-art-analysis/SKILL.md b/skills/legal/patentprior-art-analysis/SKILL.md deleted file mode 100644 index 5fb2de21..00000000 --- a/skills/legal/patentprior-art-analysis/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ ---- -name: patent-prior-art-analysis -description: Conduct comprehensive prior art searches and generate invalidation reports for patent applications and existing patents -tags: - - practice: patent - - document: analysis - - mode: research ---- - -# Patent Prior Art Analysis - -Identify and analyze prior art references that anticipate or render a patent claim obvious. - -## Prerequisites -- Patent UCID or application number -- Independent and dependent claims -- Invention disclosure or technical description -- Key technical features (one per line) - -## Workflow -1. **Search**: Run semantic search across patent and non-patent literature using technical features -2. **Filter**: Exclude references beyond the priority date and excluded citations -3. **Analyze**: For each candidate reference, assess anticipation vs. obviousness -4. **Map**: Create a claims mapping chart showing which claims each reference teaches -5. **Report**: Generate structured invalidation report with: - - Technical summaries of each reference - - Anticipation analysis per claim - - Obviousness analysis combining references - - Confidence scores for each mapping - -## Pitfalls -- Don't confuse novelty with non-obviousness; document both separately -- Always verify the reference's priority date is before the claimed invention -- Exclude references that are merely cited by the patent without being prior art -- Avoid over-indexing on keyword overlap — semantics matter more than vocabulary diff --git a/skills/legal/paxton-deposition-outline/SKILL.md b/skills/legal/paxton-deposition-outline/SKILL.md deleted file mode 100644 index 308fcc1b..00000000 --- a/skills/legal/paxton-deposition-outline/SKILL.md +++ /dev/null @@ -1,34 +0,0 @@ ---- -name: paxton-deposition-outline -tags: - practice: litigation - document: checklist - mode: research ---- - -# Deposition Examination Outline - -## Purpose -Create a structured deposition outline for direct or cross-examination of a witness. Organized by topic with sequenced questions, supporting exhibits, and fallback lines. - -## Prerequisites -- Witness name, role, and anticipated testimony area -- Case theory and key facts the witness will address (or evade) -- Prior statements, depositions, or written records from the witness -- Trial date and deposition notice (for time-boxing) - -## Workflow -1. Open with foundation questions (identity, background, qualifications) -2. Move to topics in order of strategic importance -3. For each topic: list (a) facts to establish, (b) sequenced questions, (c) supporting exhibits with exhibit numbers -4. Add fallback questions for uncooperative or evasive responses -5. Include cross-references to prior statements for impeachment -6. Close with broad invitation for the witness to add anything (direct) or last-word questions (cross) -7. Attach exhibit matrix listing each exhibit's purpose, exhibit number, and question range - -## Pitfalls -- Never ask open-ended questions on cross-examination — every question should be answerable with yes/no -- Avoid argumentative questions; they give the witness room to explain away -- Do not skip foundation — an objectionable foundation sinks the entire line -- Pre-flag potential Daubert challenges for expert witnesses -- Time-box each section and note when to cut vs. continue diff --git a/skills/legal/paxton-legal-research-memo/SKILL.md b/skills/legal/paxton-legal-research-memo/SKILL.md deleted file mode 100644 index 8fa6a064..00000000 --- a/skills/legal/paxton-legal-research-memo/SKILL.md +++ /dev/null @@ -1,33 +0,0 @@ ---- -name: paxton-legal-research-memo -tags: - practice: litigation - document: memo - mode: drafting ---- - -# Legal Research Memo Drafting - -## Purpose -Draft a comprehensive legal research memo using client-provided facts, issue framing, and jurisdictional context. Output follows standard IRAC structure with issue statement, brief answer, facts, analysis, and conclusion. - -## Prerequisites -- Client statement of facts or case brief -- Jurisdiction (court, state, federal) -- Specific legal questions or issues to research -- Any authority the client already has (cases, statutes, regulations) - -## Workflow -1. Ingest client facts and identify the legal issues presented -2. Map each issue to its elements under the relevant jurisdiction's law -3. For each element, draft a sub-analysis section applying facts to law -4. Identify counterarguments and distinguish unfavorable authority -5. Synthesize a conclusion that answers each issue directly -6. Include a "next steps" section recommending additional research, discovery, or citation verification - -## Pitfalls -- Never cite cases you cannot verify — flag any citation as "verify citation" -- Avoid conclusory analysis; every factual assertion must connect to a legal element -- Distinguish between binding and persuasive authority explicitly -- Flag statutes that may have been amended; recommend current check -- Keep tone neutral — a memo analyzes, it does not advocate diff --git a/skills/legal/paxton-policy-issue-memo/SKILL.md b/skills/legal/paxton-policy-issue-memo/SKILL.md deleted file mode 100644 index 04ced660..00000000 --- a/skills/legal/paxton-policy-issue-memo/SKILL.md +++ /dev/null @@ -1,33 +0,0 @@ ---- -name: paxton-policy-issue-memo -tags: - practice: regulatory - document: memo - mode: analysis ---- - -# Policy and Compliance Issue Memo - -## Purpose -Analyze a proposed or existing organizational policy against applicable legal requirements, identifying gaps, conflicts, and risk exposures. Output is a structured memo suitable for management review. - -## Prerequisites -- The policy text or description of the proposed policy -- Jurisdiction(s) and industry sector -- Relevant laws, regulations, and standards (e.g., SOC 2, ISO, GDPR) -- Internal risk appetite or known areas of concern - -## Workflow -1. Extract each material provision from the policy -2. Map each provision to the applicable legal requirement or standard -3. Identify provisions that: (a) fully comply, (b) partially comply with gaps, (c) conflict with law, (d) are absent but required -4. For each gap or conflict, draft a recommended revision with statutory or regulatory citation -5. Produce a risk matrix: provision → risk level (high/medium/low) → recommendation -6. Summarize top 3 priorities for remediation with implementation urgency - -## Pitfalls -- Distinguish between mandatory legal requirements and best practices — not all recommendations are legal necessities -- Flag provisions that appear to over-restrict and may create business risk -- Never state a regulation requires something unless you can cite the specific section -- When a requirement is ambiguous, flag it and propose two compliant approaches -- Keep management summary under 200 words; detailed analysis goes below diff --git a/skills/legal/permitting-diligence-report/SKILL.md b/skills/legal/permitting-diligence-report/SKILL.md deleted file mode 100644 index 15304017..00000000 --- a/skills/legal/permitting-diligence-report/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ ---- -name: permitting-diligence-report -description: Draft permitting due diligence reports analyzing approval risk, zoning compliance, and community friction -tags: - - practice: regulatory - - document: memo - - mode: analysis ---- - -# Permitting Due Diligence Report - -Produce a comprehensive permitting diligence memo for real estate developers and infrastructure investors. - -## Prerequisites -- Project details (type, size, use case, location) -- Jurisdiction (city, county, state) -- Site address or coordinates -- Capital commitment timeline and decision deadline - -## Workflow -1. **Data Capture**: Collect local government records — zoning codes, overlays, planning commission agendas, staff reports -2. **Pattern Analysis**: Compare the proposed project against past applications in the same jurisdiction -3. **Risk Assessment**: Evaluate approval risk, political friction, and community opposition patterns -4. **Memo Draft**: Generate a decision-ready memo covering: - - Zoning compliance analysis - - Historical approval/denial patterns for similar projects - - Likely conditions of approval - - Political and community risk points - - Recommended next steps and mitigation strategies - -## Pitfalls -- Don't rely solely on zoning text — staff reports and hearing transcripts often reveal de facto requirements -- Community opposition is context-specific; generic risk scores are misleading -- Moratoria and pending legislation can change the rules between research and submission -- Always distinguish between what's permitted by-right versus what requires discretionary approval diff --git a/skills/legal/planning-commission-brief/SKILL.md b/skills/legal/planning-commission-brief/SKILL.md deleted file mode 100644 index 3b0cd0e6..00000000 --- a/skills/legal/planning-commission-brief/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ ---- -name: planning-commission-brief -description: Draft planning commission presentation briefs and public hearing testimony for development projects -tags: - - practice: regulatory - - document: brief - - mode: drafting ---- - -# Planning Commission Brief - -Prepare formal briefs for planning commission hearings and public comment submissions. - -## Prerequisites -- Project description and supporting documents -- Planning commission rules and submission deadlines -- Staff report (if available) -- Community feedback and stakeholder positions - -## Workflow -1. **Structure**: Organize the brief into: - - Project summary (one page) - - Zoning and code compliance table - - Community benefits statement - - Responses to anticipated staff questions - - Supporting evidence from similar approved projects -2. **Anticipate**: Pre-address likely concerns (traffic, parking, height, shadow, character) -3. **Evidence**: Cite specific code sections and prior approvals that support the project -4. **Refine**: Keep language accessible to non-technical commissioners - -## Pitfalls -- Never argue with the staff report — frame as clarification or additional context -- Don't overwhelm with technical details; commissioners need the essentials -- Address the 3 unspoken criteria: aesthetics, precedent, and political impact -- Submit before the deadline — late briefs are often not distributed to commissioners diff --git a/skills/legal/platus-document-generation-agents/SKILL.md b/skills/legal/platus-document-generation-agents/SKILL.md deleted file mode 100644 index 1b2f949e..00000000 --- a/skills/legal/platus-document-generation-agents/SKILL.md +++ /dev/null @@ -1,36 +0,0 @@ ---- -name: platus-document-generation-agents -description: Generate legal documents at scale using AI agents with pre-stored data and template variable detection -tags: - - practice: corporate, transactional - - document: agreement, checklist - - mode: drafting ---- - -# Platus Document Generation Agents - -AI agents that auto-fill legal documents using pre-stored party data and dynamic variable detection across high-volume workflows. - -## Prerequisites - -- Document templates with defined variable fields -- Client/counterparty data store (CRM or structured data) -- Agent workflow orchestration tool -- Quality review checklist for legal accuracy - -## Workflow - -1. Load template and identify variable fields (party names, dates, amounts, jurisdiction) -2. Query stored data store for matching party information -3. Auto-populate template fields using AI variable detection -4. Run validation: check for blank fields, mismatched parties, inconsistent dates -5. Route to human reviewer for final sanity check on high-value documents -6. Generate final document set and archive with metadata - -## Pitfalls - -- Variable detection can misidentify field boundaries; always validate pre-population against template spec -- Data store may be stale — confirm party info is current before auto-generation -- Multi-party agreements need cross-verification that all parties match across documents -- Automated drafting doesn't replace legal analysis; flags are suggestions, not conclusions -- Template versioning: ensure agents use the latest approved template version diff --git a/skills/legal/platus-kyc-document-collection/SKILL.md b/skills/legal/platus-kyc-document-collection/SKILL.md deleted file mode 100644 index 5621e78e..00000000 --- a/skills/legal/platus-kyc-document-collection/SKILL.md +++ /dev/null @@ -1,37 +0,0 @@ ---- -name: platus-kyc-document-collection -description: Automate KYC document collection, verification, and compliance workflows for legal transactions and client onboarding -tags: - - practice: regulatory, corporate - - document: checklist, policy - - mode: drafting, analysis ---- - -# Platus KYC Document Collection - -Automated KYC document collection and verification for client onboarding and transactional compliance. - -## Prerequisites - -- KYC checklist template with jurisdiction-specific requirements -- Secure document intake portal (API or web link) -- Identity verification data source (government databases, third-party KYC providers) -- Compliance policy reference for target jurisdiction - -## Workflow - -1. Define KYC requirements based on transaction type and jurisdiction -2. Generate client intake portal link with required document checklist -3. Trigger automated collection: clients upload ID, proof of address, corporate docs -4. Validate uploaded documents against checklist (format, completeness, expiration) -5. Cross-reference identity data with verification sources -6. Flag discrepancies or missing documents for manual review -7. Generate compliance summary report with pass/fail status - -## Pitfalls - -- KYC requirements differ significantly by jurisdiction — don't apply one-size-fits-all checklists -- Document expiration dates matter; expired IDs fail verification even if content is correct -- Corporate structures (LLC, trust, foundation) require layered documentation — entity docs + beneficial owner info -- Privacy laws (GDPR, CCPA) restrict how identity data is stored and processed -- Automated validation catches format issues but not sophisticated document forgery diff --git a/skills/legal/platus-notarization-api/SKILL.md b/skills/legal/platus-notarization-api/SKILL.md deleted file mode 100644 index 8032c0d5..00000000 --- a/skills/legal/platus-notarization-api/SKILL.md +++ /dev/null @@ -1,36 +0,0 @@ ---- -name: platus-notarization-api -description: Automate high-volume document notarization and electronic signatures at scale using API-driven workflows -tags: - - practice: transactional - - document: agreement, letter - - mode: drafting, analysis ---- - -# Platus Notarization & Signature API - -Automate batch notarization and electronic signature collection for legal documents. Ideal for POAs, NDAs, and agreements processed in volume. - -## Prerequisites - -- Access to notarization API endpoint (Platus or equivalent) -- Document templates with variable placeholders -- Agent identity credentials for API auth -- Understanding of e-signature legal requirements (ESIGN/UETA) - -## Workflow - -1. Identify documents requiring notarization or signatures -2. Populate templates using stored client data and extracted variables -3. Route documents through API endpoint for electronic signing -4. Trigger notarization workflow for documents requiring notary attestation -5. Collect signed/notarized documents and return confirmation metadata -6. Log execution results: timestamps, signer IDs, completion status - -## Pitfalls - -- Notarization requirements vary by jurisdiction and document type; verify remotely online notarization (RON) eligibility -- ESIGN compliance requires explicit consent and audit trails — don't skip consent capture -- API rate limits on batch submissions; throttle for >1000 document volumes -- Some jurisdictions still require wet signatures for real estate deeds, marriage docs, wills -- Verify that your notarization provider meets state-specific RON requirements diff --git a/skills/legal/pretorin-fedramp-compliance-drafting/SKILL.md b/skills/legal/pretorin-fedramp-compliance-drafting/SKILL.md deleted file mode 100644 index 06db3271..00000000 --- a/skills/legal/pretorin-fedramp-compliance-drafting/SKILL.md +++ /dev/null @@ -1,31 +0,0 @@ -# FedRAMP Security Plan & Control Implementation Plan Drafting - -## Prerequisites -- Target system architecture and deployment model (cloud/on-prem/hybrid) -- Existing security controls inventory or assessment results -- FedRAMP baseline level (Low/Moderate/High) -- Agency sponsor requirements (if applicable) - -## Workflow -1. Determine FedRAMP baseline level from system categorization (FIPS 199/53) -2. Map applicable controls from NIST SP 800-53 to FedRAMP control catalog -3. For each control, draft the Security Plan section covering: - - Control intent and implementation approach - - Responsible parties (roles: CSP, AO, ISSO, CISO) - - Implementation artifacts (policies, procedures, technical configs) - - Inherited vs. shared vs. individual control designation -4. Generate the Control Implementation Plan (CIP) with implementation timeline -5. Flag gaps where controls lack documentation or technical evidence -6. Output structured FedRAMP Security Plan sections ready for human review - -## Pitfalls -- Controls are inherited/shared/individual — misclassification causes AO rejection -- SP 800-53 Rev 5 has different numbering than Rev 4 — ensure version alignment -- FedRAMP requires family-specific implementation notes (AC, AU, CA, CM, etc.) -- Don't draft controls that are outside the system boundary -- Always reference current FedRAMP Control Baselines, not deprecated versions - -## Tags -practice_area: regulatory -document_type: policy, checklist -skill_mode: drafting diff --git a/skills/legal/pretorin-nist-control-mapping/SKILL.md b/skills/legal/pretorin-nist-control-mapping/SKILL.md deleted file mode 100644 index 46e75a9a..00000000 --- a/skills/legal/pretorin-nist-control-mapping/SKILL.md +++ /dev/null @@ -1,31 +0,0 @@ -# Cross-Framework Security Control Mapping & Gap Analysis - -## Prerequisites -- List of applicable frameworks (FedRAMP, DoD RMF, NIST 800-53, CMMC, ISO 27001) -- Current system security posture documentation -- Existing control evidence or assessment reports - -## Workflow -1. Identify all applicable frameworks and their current baseline versions -2. For each security control in the target framework: - - Find equivalent controls across all other frameworks - - Build a cross-reference mapping table (control ID → equivalents) - - Flag controls unique to one framework that lack cross-walks -3. Assess implementation status for each mapped group: - - Fully implemented (all frameworks satisfied) - - Partially implemented (some frameworks have evidence, others need work) - - Not implemented (gap — needs remediation plan) -4. Generate a prioritized remediation backlog ranked by framework dependency -5. Output a matrix showing control coverage across all frameworks - -## Pitfalls -- DoD RMF (DFARS 7012) has additional controls beyond standard NIST 800-53 -- CMMC 2.0 maps to NIST 800-171 (not 800-53) — don't conflate the two -- Control enhancements (e.g., AC-2(3)) have different numbering — preserve full IDs -- Some frameworks require supplementary guidance documents (RMF Guide, CNSS 1253) -- CMMC third-party assessments vs. self-assessments have different control sets - -## Tags -practice_area: regulatory -document_type: checklist, analysis -skill_mode: analysis diff --git a/skills/legal/pretorin-rmf-evidence-organization/SKILL.md b/skills/legal/pretorin-rmf-evidence-organization/SKILL.md deleted file mode 100644 index d33ff14b..00000000 --- a/skills/legal/pretorin-rmf-evidence-organization/SKILL.md +++ /dev/null @@ -1,34 +0,0 @@ -# RMF Authorization Package Evidence Package Organization - -## Prerequisites -- System Security Plan (SSP) or draft -- List of all controls assigned to the system -- Available evidence sources (config snapshots, logs, policy docs, test results) - -## Workflow -1. Parse the SSP to extract each control and its implementation statement -2. For every control in the package (SSP, SARA, SAR, PP&RL, Plan of Actions): - - Identify required evidence artifacts per control family - - Cross-reference against available evidence inventory - - Mark as: present, partial, missing, or stale -3. Organize evidence into the standard RMF package structure: - - SSP — system description, boundary diagram, data flow - - SARA — security assessment results with test evidence - - POA&M — remediation plan for findings - - SAP — authorizing official decision package - - Related plans (PP&RL, contingency, continuity) -4. Generate a traceability matrix: control → evidence → assessor notes -5. Highlight any controls with insufficient evidence for the planned assessment type -6. Output the organized evidence index ready for human verification - -## Pitfalls -- Evidence must be current — snapshots older than 90 days need refreshing -- Continuous monitoring evidence (SCUM, automated scanning) is now expected -- SARA requires actual test procedures, not just policy references -- Authority to Operate (ATO) packages differ by AO — know the sponsor's format -- POA&M entries need realistic milestones and realistic remediation dates - -## Tags -practice_area: regulatory -document_type: checklist -skill_mode: analysis diff --git a/skills/legal/project-approval-risk-assessment/SKILL.md b/skills/legal/project-approval-risk-assessment/SKILL.md deleted file mode 100644 index 936cd9b2..00000000 --- a/skills/legal/project-approval-risk-assessment/SKILL.md +++ /dev/null @@ -1,34 +0,0 @@ ---- -name: project-approval-risk-assessment -description: Assess approval risk for development projects based on historical data, community patterns, and political context -tags: - - practice: regulatory - - document: summary - - mode: analysis ---- - -# Project Approval Risk Assessment - -Evaluate the likelihood of approval for a development project using historical precedent and political analysis. - -## Prerequisites -- Project type and scale -- Jurisdiction with planning commission history -- Project timeline and submission target -- Known community stakeholders or opponents - -## Workflow -1. **Historical Research**: Analyze past applications, hearing transcripts, and outcomes for similar projects -2. **Stakeholder Mapping**: Identify likely supporters and opponents from community records -3. **Pattern Analysis**: Determine common approval/denial reasons for comparable projects -4. **Risk Scoring**: Assign quantitative risk levels: - - High risk: frequent denials, strong organized opposition - - Medium risk: conditional approvals common, delays likely - - Low risk: consistent approvals with standard conditions -5. **Report**: Generate risk memo with confidence levels and recommended strategies - -## Pitfalls -- Past decisions don't guarantee future outcomes — political winds shift -- Staff recommendations carry weight but aren't binding — note this distinction -- Community opposition that lost before may mobilize around a new project -- Don't confuse procedural delays with substantive denial risk diff --git a/skills/legal/regulatory-controls-design/SKILL.md b/skills/legal/regulatory-controls-design/SKILL.md deleted file mode 100644 index 1f91a7d1..00000000 --- a/skills/legal/regulatory-controls-design/SKILL.md +++ /dev/null @@ -1,38 +0,0 @@ -name: regulatory-controls-design -language: en -description: Generates context-aware compliance controls tailored to an organization's specific profile — company size, industry, data handling practices, and risk tolerance. Produces implementable control specifications aligned to multiple frameworks simultaneously. -tags: - - drafting - - checklist - - regulatory ---- - -# AI-Generated Controls Design - -Creates context-aware compliance controls and policy specifications tailored to an organization's unique profile, derived from regulatory requirements and mapped to applicable frameworks. - -## Prerequisites - -Before executing, collect: - -1. **Organization profile** — industry, size, data types processed, cloud infrastructure -2. **Target regulations** — which laws/frameworks apply (GDPR, SOC 2, NIST, ISO 27001, etc.) -3. **Existing controls inventory** — current policies, technical controls, organizational measures -4. **Risk appetite** — conservative (zero tolerance), moderate, or risk-tolerant approach - -## Workflow - -1. **Regulatory requirement extraction** — parse applicable requirements from target frameworks -2. **Contextual analysis** — map requirements to organization's actual infrastructure and practices -3. **Control specification generation** — draft specific, measurable control requirements with implementation guidance -4. **Cross-framework alignment** — identify controls that satisfy multiple framework requirements simultaneously -5. **Role assignment** — assign control owners and review responsibilities based on org structure -6. **Validation checklist** — produce evidence checklist for auditors to verify control implementation -7. **Periodic refresh scheduling** — set review cadence based on regulatory change frequency - -## Pitfalls - -- Generic controls fail audits — must be specific to the organization's actual infrastructure -- Overlap between frameworks can create false sense of compliance — map carefully -- Technical controls must match actual tech stack — don't prescribe unsupported technology -- Human-facing controls (training, policies) need change management, not just documentation diff --git a/skills/legal/settleindex-casebot-risk-model/SKILL.md b/skills/legal/settleindex-casebot-risk-model/SKILL.md deleted file mode 100644 index 9b277dd5..00000000 --- a/skills/legal/settleindex-casebot-risk-model/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ ---- -name: casebot-risk-model -description: Auto-generate dispute risk models from pleadings — the fastest system for modeling complex litigation without manual setup -tags: - - practice_area: litigation - - document_type: analysis, checklist - - skill_mode: analysis ---- - -# CaseBot Risk Model Generator - -Automated case modeling from pleadings — extract claims, defenses, damages, and key legal issues without manual configuration. - -## Prerequisites - -- Court pleadings (any format: PDF, Word, text) -- Jurisdiction identification (auto-detected or specified) -- Calibration data for the relevant court/tribunal (optional but recommended) - -## Workflow - -1. Ingest pleadings in any format — automatic parsing and entity extraction -2. CaseBot identifies: claimant, defendant, cause(s) of action, damages sought, key defenses -3. Maps extracted data against jurisdiction-specific risk calibration databases -4. Generates a structured risk model with weighted probabilities per claim/defense -5. Outputs a concise risk summary with key factors driving the model -6. Supports API-based integration for enterprise platform embedding -7. Enables SSO authentication for team-based access and audit trails - -## Pitfalls - -- Pleadings must contain sufficient factual detail for the model to calibrate -- Automatic jurisdiction detection should be verified — misidentified courts produce wrong weights -- The model is a tool for strategic insight, not a substitute for legal analysis -- Always review the extracted claims/defenses against the original pleadings before relying on the model diff --git a/skills/legal/settleindex-litigation-risk-modeling/SKILL.md b/skills/legal/settleindex-litigation-risk-modeling/SKILL.md deleted file mode 100644 index 74db1366..00000000 --- a/skills/legal/settleindex-litigation-risk-modeling/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ ---- -name: litigation-risk-modeling -description: Build automated litigation risk models from pleadings, calculate probability-weighted outcomes for claimant and defendant perspectives -tags: - - practice_area: litigation - - document_type: analysis, memo - - skill_mode: analysis ---- - -# Automated Litigation Risk Modeling - -Generate deterministic risk models from legal pleadings with dual-party perspective analysis and probability-weighted outcomes. - -## Prerequisites - -- Court pleadings (statement of claim, defense, reply) -- Applicable jurisdiction's case law database -- Cost framework for the relevant court/tribunal - -## Workflow - -1. Ingest pleadings and automatically extract the dispute model (parties, claims, defenses, damages sought) -2. Build a risk model using jurisdiction-specific precedent weights -3. Generate claimant-side outcome probability distribution (win/loss/ partial, expected damages) -4. Generate defendant-side outcome probability distribution independently -5. Identify the Model Settlement Value where both probability lines intersect -6. Apply lawyer refinement overlay for strategy, costs, and settlement offers -7. Output a Settle Chart showing both parties' risk curves and the Model Settlement Value - -## Pitfalls - -- Risk models are probabilistic, not deterministic — always present confidence ranges -- Lawyer refinement must be clearly labeled as advisory, not replacing legal judgment -- Cost projections must be jurisdiction-specific; generic cost models produce misleading results -- Dual-party models must be generated independently to avoid confirmation bias diff --git a/skills/legal/settleindex-settlement-value-forecast/SKILL.md b/skills/legal/settleindex-settlement-value-forecast/SKILL.md deleted file mode 100644 index dd5e98a7..00000000 --- a/skills/legal/settleindex-settlement-value-forecast/SKILL.md +++ /dev/null @@ -1,35 +0,0 @@ ---- -name: settlement-value-forecasting -description: Predict precise settlement values from existing claim documents using AI-powered modeling without manual data entry -tags: - - practice_area: litigation - - document_type: analysis, summary - - skill_mode: analysis ---- - -# Settlement Value Forecasting - -Calculate settlement values and expected outcomes directly from claim documents using automated AI case modeling. - -## Prerequisites - -- Claim documents (complaints, statements of case, witness statements) -- Defendant response documents (if filed) -- Relevant jurisdiction and court level - -## Workflow - -1. Upload claim documents — no manual data entry required -2. Casebot automatically builds the dispute model from document text in under one minute -3. Run dual-party risk analysis: calculate what claimant and defendant would each value the case at -4. Generate the Model Settlement Value from the intersection of both parties' cost-spend curves -5. Produce a Settle Chart visualization showing current settlement ranges -6. Support portfolio-level analytics when multiple cases are analyzed together -7. Output is audit-ready with deterministic, repeatable methodology - -## Pitfalls - -- Model Settlement Value depends on accurate cost data — verify cost assumptions -- Portfolio analytics require consistent case classification across all matters -- The model counters bias but cannot correct for fundamentally flawed input documents -- Always validate the Casebot's extracted model against the actual pleadings diff --git a/skills/legal/solomon-tax-preparation-workflow/SKILL.md b/skills/legal/solomon-tax-preparation-workflow/SKILL.md deleted file mode 100644 index 459786ad..00000000 --- a/skills/legal/solomon-tax-preparation-workflow/SKILL.md +++ /dev/null @@ -1,69 +0,0 @@ ---- -name: tax-preparation-workflow -language: en -description: >- - Automates tax preparation workflows from document collection through return - generation. Handles income verification, deduction identification, form - completion, and compliance checking for individual and business returns. - Use when preparing tax returns, organizing tax documents, or ensuring - compliance across multiple filing periods. -tags: - - compliance - - regulatory - - drafting ---- - -# Tax Preparation Workflow - -Automated workflow for end-to-end tax preparation: from document ingestion -through return generation and compliance verification. - -## Prerequisites - -- Tax documents (W-2s, 1099s, K-1s, receipts, prior-year returns) -- Filing status and dependency information -- Business entity documentation (if applicable) -- Tax year and jurisdiction(s) - -## Workflow - -1. **Document collection and categorization** — ingest and classify all tax documents: - - Income documents (W-2, 1099 series, K-1, 1098, statements) - - Deduction evidence (receipts, invoices, mileage logs, charitable contributions) - - Credits documentation (education, energy, childcare, adoption) - - Prior-year returns (for carryforward items: NOL, credits, losses) -2. **Income reconstruction** — build complete income picture: - - Aggregate all income sources by type (wage, self-employment, investment, passive) - - Identify unreported income from information returns - - Reconcile with bank statements and financial records - - Flag discrepancies between reported and claimed income -3. **Deduction and credit identification** — map available deductions and credits: - - Standard vs. itemized deduction analysis - - Above-the-line deductions (IRA, student loan, HSA) - - Schedule C deductions (self-employment) - - Credit optimization (EITC, CTC, AOTC, Saver's Credit) - - State-specific deductions and credits -4. **Form preparation and assembly** — generate all required forms: - - Core forms (1040, schedules, attachments) - - Entity forms (1120, 1120-S, 1065, 941) - - Information returns (1099, 1096, W-2c) - - State and local returns -5. **Compliance verification** — validate the prepared return: - - Mathematical accuracy check - - Tax law compliance (recent changes, updates, expiration of provisions) - - Consistency check across all forms and schedules - - Prior-year carryforward reconciliation -6. **Review and filing** — produce final deliverables: - - Draft return with supporting worksheets - - Tax liability/schedule with cash flow impact - - Filing instructions and deadline calendar - - Recommended advisory items (estimated payments, planning opportunities) - -## Pitfalls - -- Information returns (1099s) filed by payers must match the taxpayer's return — mismatch triggers IRS correspondence -- Standard deduction amounts change annually — verify the correct year's threshold -- Tax law provisions have expiration dates — a deduction available this year may not exist next year -- State filing requirements can be triggered by remote activities — don't assume one state covers all -- Estimated tax penalties apply when payments fall below 90% of current-year or 100% of prior-year tax — calculate exposure proactively -- Carryforward items (NOL, capital losses, tax credits) from prior years are the most common source of filing errors — always reconcile diff --git a/skills/legal/trustplane-compliance-audit/SKILL.md b/skills/legal/trustplane-compliance-audit/SKILL.md deleted file mode 100644 index c4503ab9..00000000 --- a/skills/legal/trustplane-compliance-audit/SKILL.md +++ /dev/null @@ -1,25 +0,0 @@ ---- -name: trustplane-compliance-audit -description: Generate audit-ready compliance reports for LLM deployments covering EU AI Act, NIST AI RMF, SOC 1, HIPAA, and GDPR requirements. -tags: [compliance, regulatory] -skill_mode: [drafting, analysis] ---- - -# LLM Compliance Audit Report Generation - -## Prerequisites -- Audit logs of all LLM interactions (prompts, completions, enforcement actions) -- Registry of deployed models and their risk classifications -- Current regulatory framework requirements - -## Workflow -1. Compile the AI inventory: all models, use cases, data sources, and ownership -2. Extract enforcement data from audit logs: block rates, warning rates, attack patterns -3. Map compliance evidence to each framework requirement (EU AI Act articles, NIST functions, SOC controls) -4. Generate narrative analysis explaining the risk posture and control effectiveness -5. Produce framework-specific report sections with traceable evidence references - -## Pitfalls -- Never extrapolate beyond logged data — if it's not in the audit log, it's not proven -- Separate compliance evidence from compliance claims; only report verified controls -- Update reports whenever model configurations or data sources change diff --git a/skills/legal/trustplane-llm-safety/SKILL.md b/skills/legal/trustplane-llm-safety/SKILL.md deleted file mode 100644 index f1b8a287..00000000 --- a/skills/legal/trustplane-llm-safety/SKILL.md +++ /dev/null @@ -1,24 +0,0 @@ ---- -name: trustplane-llm-safety -description: Draft safety policies and prompt inspection rules for LLM deployments — detect PII, prompt injection, impersonation, and sensitive data exposure. -tags: [compliance, regulatory] -skill_mode: [drafting, analysis] ---- - -# LLM Safety Policy & Prompt Inspection Rules - -## Prerequisites -- Knowledge of the LLM use case and data sensitivity classification -- Applicable regulatory frameworks (EU AI Act, GDPR, HIPAA, etc.) - -## Workflow -1. Classify the LLM deployment by risk tier (minimal, limited, high, unacceptable) -2. Define forbidden input categories: PII, PHI, confidential business data, prompt injections -3. Draft threshold-based detection rules for each risk category with confidence scoring -4. Specify enforcement actions per category: ALLOW, WARN, BLOCK -5. Design audit logging requirements capturing every prompt, completion, and enforcement decision - -## Pitfalls -- Don't block legitimate business queries with overly broad PII rules — tune thresholds -- Ensure prompt-hacking detection covers multi-turn and indirect injection patterns -- Log all enforcement decisions with reason codes for regulator review diff --git a/skills/legal/trustplane-prompt-inspection/SKILL.md b/skills/legal/trustplane-prompt-inspection/SKILL.md deleted file mode 100644 index edecc360..00000000 --- a/skills/legal/trustplane-prompt-inspection/SKILL.md +++ /dev/null @@ -1,24 +0,0 @@ ---- -name: trustplane-prompt-inspection -description: Review and classify LLM prompts against safety, compliance, and data governance policies — categorize risk, flag policy violations, and recommend actions. -tags: [compliance, regulatory] -skill_mode: [analysis, research] ---- - -# Prompt Classification & Risk Analysis - -## Prerequisites -- Published safety policy with defined risk categories and thresholds -- Access to prompt content and associated model configuration - -## Workflow -1. Parse the prompt for content categories: PII, PHI, sensitive data, prompt injection, impersonation -2. Score each category on a 0-1 confidence scale against defined thresholds -3. Classify the overall prompt: clean, warn, or block based on combined scores -4. Document the classification rationale with category scores and matched rules -5. Recommend action: pass through, add warning, or block with reason - -## Pitfalls -- Don't classify a prompt solely on keyword matching — context matters (e.g., SSN in a redacted filing vs. raw data) -- Distinguish between prompt-hacking intent and accidental policy violations -- Maintain consistent scoring across all categories for audit comparability diff --git a/skills/legal/zoning-compliance-analysis/SKILL.md b/skills/legal/zoning-compliance-analysis/SKILL.md deleted file mode 100644 index 5761393b..00000000 --- a/skills/legal/zoning-compliance-analysis/SKILL.md +++ /dev/null @@ -1,31 +0,0 @@ ---- -name: zoning-compliance-analysis -description: Analyze proposed projects against local zoning codes and overlay districts for compliance issues -tags: - - practice: regulatory - - document: analysis - - mode: analysis ---- - -# Zoning Compliance Analysis - -Evaluate whether a proposed project complies with applicable zoning ordinances and overlay districts. - -## Prerequisites -- Proposed use case, building dimensions, and site layout -- Address or APN (Assessor's Parcel Number) -- Current zoning designation and any overlay districts -- Planned development or conditional use permit status - -## Workflow -1. **Research**: Pull applicable zoning code sections, overlay regulations, and local amendments -2. **Map Requirements**: Extract numeric requirements (height, FAR, setbacks, parking, landscaping) -3. **Compare**: Measure proposed project against each requirement -4. **Flag Issues**: Identify where the project exceeds limits or falls short -5. **Recommend**: Suggest variances, conditional use permits, or design modifications - -## Pitfalls -- Overlay districts often impose additional restrictions beyond base zoning -- Grandfathered uses may be non-conforming — verify current status -- Parking requirements can be location-specific (transit, bike share, EV) -- Temporary uses (events, sales offices) often have separate code sections