Add technical-change-tracker skill

.githooks/pre-commit

@@ -0,0 +1,53 @@
+#!/bin/bash
+# Pre-commit hook: scan staged changes for sensitive data
+# Install: git config core.hooksPath .githooks
+
+set -euo pipefail
+
+RED='\033[0;31m'
+GREEN='\033[0;32m'
+YELLOW='\033[1;33m'
+NC='\033[0m'
+
+echo "🔍 Scanning staged changes for sensitive data..."
+
+FAILED=0
+
+# Layer 1: gitleaks (if available)
+if command -v gitleaks &>/dev/null; then
+  if ! gitleaks protect --staged --config .gitleaks.toml --no-banner 2>/dev/null; then
+    echo -e "${RED}❌ gitleaks found secrets in staged changes${NC}"
+    FAILED=1
+  fi
+else
+  echo -e "${YELLOW}⚠ gitleaks not installed (brew install gitleaks), falling back to pattern scan${NC}"
+fi
+
+# Layer 2: fast regex scan (always runs, catches what gitleaks config might miss)
+STAGED_DIFF=$(git diff --cached --diff-filter=ACDMR)
+
+PATTERNS=(
+  '/Users/[a-zA-Z][a-zA-Z0-9_-]+/'
+  '/home/[a-zA-Z][a-zA-Z0-9_-]+/'
+  'C:\\Users\\[a-zA-Z]'
+  'songtiansheng'
+  'tiansheng'
+  '15366[0-9]+'
+)
+
+for pattern in "${PATTERNS[@]}"; do
+  MATCHES=$(echo "$STAGED_DIFF" | grep -nE "^\+" | grep -E "$pattern" | grep -v "^+++\|\.gitleaks\.toml\|\.githooks/\|\.gitignore\|placeholder\|example\|CLAUDE\.md" || true)
+  if [ -n "$MATCHES" ]; then
+    echo -e "${RED}❌ Found sensitive pattern '${pattern}':${NC}"
+    echo "$MATCHES" | head -5
+    FAILED=1
+  fi
+done
+
+if [ $FAILED -eq 1 ]; then
+  echo ""
+  echo -e "${RED}Commit blocked. Fix the issues above, or use --no-verify to bypass (not recommended).${NC}"
+  exit 1
+fi
+
+echo -e "${GREEN}✅ No sensitive data found in staged changes.${NC}"

.github/ISSUE_TEMPLATE/bug_report.md

@@ -16,7 +16,7 @@ Which skill is affected?
 
 - [ ] skill-creator
 - [ ] github-ops
-- [ ] markdown-tools
+- [ ] doc-to-markdown
 - [ ] mermaid-tools
 - [ ] statusline-generator
 - [ ] teams-channel-post-writer

.github/ISSUE_TEMPLATE/feature_request.md

@@ -20,7 +20,7 @@ Which skill would this enhance?
 
 - [ ] skill-creator
 - [ ] github-ops
-- [ ] markdown-tools
+- [ ] doc-to-markdown
 - [ ] mermaid-tools
 - [ ] statusline-generator
 - [ ] teams-channel-post-writer

.github/PULL_REQUEST_TEMPLATE.md

@@ -33,7 +33,7 @@ Which skills are affected by this PR?
 
 - [ ] skill-creator
 - [ ] github-ops
-- [ ] markdown-tools
+- [ ] doc-to-markdown
 - [ ] mermaid-tools
 - [ ] statusline-generator
 - [ ] teams-channel-post-writer

.gitignore

@@ -59,6 +59,7 @@ Thumbs.db
 *.tgz
 *.rar
 *.7z
+*.skill
 
 # Build artifacts
 *.o
@@ -81,3 +82,14 @@ INSTALLATION.md
 # Private/commercial skills (moved to claude-code-skills-pro)
 seo-expert/
 video-creator/
+/jsonl-viewer/
+
+# Research output (may contain sensitive data)
+deep-research-output/
+recovered_deep_research/
+
+# OpenCLI cache
+.opencli/
+
+# Eval workspaces (contain test data with personal info)
+douban-skill-workspace/

.gitleaks.toml

@@ -0,0 +1,53 @@
+# Gitleaks custom rules for claude-code-skills repo
+# Catches personal info that shouldn't be in an open source repo
+
+title = "claude-code-skills sensitive data rules"
+
+[extend]
+useDefault = true
+
+# Global allowlist: files that are allowed to contain patterns
+# (the config file itself, hooks, and contribution guides)
+[allowlist]
+paths = [
+  '''\.gitleaks\.toml$''',
+  '''\.githooks/''',
+  '''CONTRIBUTING\.md$''',
+  '''CLAUDE\.md$''',
+]
+
+[[rules]]
+id = "absolute-user-path-macos"
+description = "Hardcoded macOS user home directory path"
+regex = '''/Users/[a-zA-Z][a-zA-Z0-9_-]+/'''
+tags = ["pii", "path"]
+
+[[rules]]
+id = "absolute-user-path-linux"
+description = "Hardcoded Linux home directory path"
+regex = '''/home/[a-zA-Z][a-zA-Z0-9_-]+/'''
+tags = ["pii", "path"]
+
+[[rules]]
+id = "windows-user-path"
+description = "Hardcoded Windows user profile path"
+regex = '''C:\\Users\\[a-zA-Z][a-zA-Z0-9_-]+\\'''
+tags = ["pii", "path"]
+
+[[rules]]
+id = "phone-number-cn"
+description = "Chinese mobile phone number"
+regex = '''1[3-9]\d{9}'''
+tags = ["pii", "phone"]
+
+[[rules]]
+id = "douban-user-id-literal"
+description = "Hardcoded Douban user ID"
+regex = '''songtiansheng'''
+tags = ["pii", "username"]
+
+[[rules]]
+id = "email-personal"
+description = "Personal email address"
+regex = '''[a-zA-Z0-9._%+-]+@(gmail|qq|163|126|outlook|hotmail|yahoo|icloud|foxmail)\.[a-zA-Z]{2,}'''
+tags = ["pii", "email"]

CHANGELOG.md

@@ -7,8 +7,54 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+### Changed
+- **Renamed**: `markdown-tools` → `doc-to-markdown` — clearer name for DOCX/PDF/PPTX → Markdown conversion
+- **doc-to-markdown**: Added 8 DOCX post-processing fixes (grid tables, simple tables, CJK bold spacing, JSON pretty-print, image path flattening, pandoc attribute cleanup, code block detection, bracket fixes)
+- **doc-to-markdown**: Added 31 unit tests (`test_convert.py`)
+- **doc-to-markdown**: Added 5-tool benchmark report (`references/benchmark-2026-03-22.md`)
+
+## [1.39.0] - 2026-03-18
+
 ### Added
-- None
+- **New Skill**: scrapling-skill v1.0.0 - Reliable Scrapling CLI installation, troubleshooting, and extraction workflows for HTML, Markdown, and text output
+  - Bundled `diagnose_scrapling.py` script to verify CLI health, detect missing extras, inspect Playwright browser runtime, and run real smoke tests
+  - Static-first workflow for choosing between `extract get`, `extract fetch`, and `stealthy-fetch`
+  - Verified WeChat public article extraction pattern using `#js_content`
+  - Verified recovery path for local TLS trust-store failures via `--no-verify`
+  - Bundled troubleshooting reference covering extras, browser runtime, and output validation
+
+### Changed
+- **skill-creator** v1.5.0 → v1.5.1: Fixed `scripts/package_skill.py` so it works when invoked directly from the repository root instead of only via `python -m`
+- **continue-claude-work** v1.1.0 → v1.1.1: Replaced newer Python-only type syntax in `extract_resume_context.py` so the script runs under the local `python3` environment
+- Updated marketplace skills/plugins count from 42 to 43
+- Updated marketplace version from 1.38.0 to 1.39.0
+- Updated marketplace metadata description to include Scrapling CLI extraction workflows
+- Updated README.md and README.zh-CN.md badges, installation commands, skill listings, use cases, quick links, and requirements
+- Updated CLAUDE.md counts, version reference, and Available Skills list (added #43)
+
+## [1.38.0] - 2026-03-07
+
+### Added
+- **New Skill**: continue-claude-work v1.1.0 - Recover local `.claude` session context and continue interrupted work without `claude --resume`
+  - Bundled Python script (`extract_resume_context.py`) for one-call context extraction
+  - Compact-boundary-aware extraction using `isCompactSummary` flag (highest-signal context from session compaction summaries)
+  - Subagent workflow recovery — parses `subagents/` directory to report completed vs interrupted agents with last outputs
+  - Session end reason detection — classifies clean exit, interrupted (ctrl-c), error cascade, or abandoned
+  - Size-adaptive reading strategy based on file size and compaction count
+  - Noise filtering — skips progress/queue-operation/api_error (37-53% of session lines)
+  - Self-session exclusion, stale index fallback, ghost session warnings
+  - MEMORY.md and session-memory integration, git workspace state fusion
+
+### Changed
+- **skill-creator** v1.4.1 → v1.5.0: SKILL.md rewrite, added eval benchmarking system (run_eval, run_loop, aggregate_benchmark), agents (analyzer, comparator, grader), eval-viewer, and improve_description script
+- **transcript-fixer** v1.1.0 → v1.2.0: `--domain` defaults to all domains, added `get_domain_stats()`, cross-domain listing, and zero-match hints
+- **tunnel-doctor** v1.3.0 → v1.4.0: Added Step 2C-1 for local vanity domain proxy interception, bundled `quick_diagnose.py` automated diagnostic script
+- **pdf-creator** v1.0.0 → v1.1.0: Replaced Python `markdown` library with pandoc for MD→HTML conversion, removed `_ensure_list_spacing` workaround
+- **github-contributor** v1.0.2 → v1.0.3: Fixed gh CLI field name (`stargazersCount` → `stargazerCount`), added Prerequisites section
+- Updated marketplace skills/plugins count from 41 to 42
+- Updated marketplace version from 1.37.0 to 1.38.0
+- Updated README.md and README.zh-CN.md badges, installation commands, skill listings, use cases, quick links, and requirements
+- Updated CLAUDE.md counts, version reference, and Available Skills list (added #42)
 
 ## [1.37.0] - 2026-03-02