feat(automation): continual learning, agents-md-audit, shell entry points#474
Merged
Conversation
…entry points Wire continual-learning skill and agents-memory-updater subagent, add a read-only AGENTS.md audit agent, shell/Makefile shortcuts, and docs for the expanded repository inspection automation stack. Co-authored-by: Bryan
Contributor
|
@Bryan-Roe Thanks for sending me some feedback. Unfortunately, I hit an error while trying to use the custom Copilot setup steps configured for this repository. The error I am seeing is: Once you or someone with the necessary access fixes the problem, please let me know in a comment and I'll try again. Thanks! |
Contributor
Reviewer's GuideImplements a new AGENTS.md audit automation agent, wires it into the existing repo automation workflow (including Make/ shell entrypoints), and introduces a Cursor-based continual-learning workflow that keeps AGENTS.md Learned sections in sync with mined transcripts. Sequence diagram for continual-learning skill and agents-memory-updater workflowsequenceDiagram
actor Developer
participant CursorChat as CursorChat
participant ContinualLearning as continual-learning
participant MemoryUpdater as agents-memory-updater
participant AgentsMd as AGENTS.md
participant Index as continual-learning-index.json
Developer->>CursorChat: invoke /continual-learning
CursorChat->>ContinualLearning: trigger SKILL.md workflow
ContinualLearning->>MemoryUpdater: call agents-memory-updater
MemoryUpdater->>AgentsMd: read AGENTS.md
MemoryUpdater->>Index: read .cursor/hooks/state/continual-learning-index.json
MemoryUpdater->>MemoryUpdater: mine transcript deltas
MemoryUpdater->>AgentsMd: update Learned sections (<=12 bullets)
MemoryUpdater->>Index: write updated incremental index
MemoryUpdater-->>ContinualLearning: updater result
ContinualLearning-->>CursorChat: return result
CursorChat-->>Developer: show updated memory status or "No high-signal memory updates."
Flow diagram for repo automation entrypoints and agents-md-audit agentflowchart LR
Dev[Developer] -->|make agents / make agents-dry| Make
Dev -->|./scripts/run_ai_automation.sh| Shell
Dev -->|python scripts/run_repo_agents.py| Runner
Make --> Runner
Shell --> Runner
Runner[scripts/run_repo_agents.py]
subgraph Agents
A1[status-freshness]
A2[marker-audit]
A3[docstring-audit]
A4[agents-md-audit]
end
Runner --> A1
Runner --> A2
Runner --> A3
Runner --> A4
A4 --> AGENTS[AGENTS.md Learned sections]
File-Level Changes
Tips and commandsInteracting with Sourcery
Customizing Your ExperienceAccess your dashboard to:
Getting Help
|
github-actions
Bot
added
documentation
🟡 Coverage —
|
| Metric | Value |
|---|---|
| Total coverage | 65.2% |
▲ vs main |
+0.1% |
| Minimum threshold | 60% |
Updated on every push · 2026-06-21
Contributor
There was a problem hiding this comment.
Pull request overview
This PR extends the repository's deterministic automation-agent framework (scripts/agents/) with a new read-only agents-md-audit agent that validates the structure and hygiene of AGENTS.md Learned sections (required sections, bullet limits, merge-conflict markers, secret patterns, and stale dates). It also adds developer entry points (make agents/agents-dry/ai-automation, plus a scripts/run_ai_automation.sh wrapper), wires the agent into the shared run_repo_agents.py orchestration, introduces a Cursor-native continual-learning skill + subagent for keeping AGENTS.md in sync, and corrects documentation (notably the --run-agents ordering, which the code runs after the integration contract gate).
Changes:
- New
agents-md-auditautomation agent + dedicated test suite, registered inrun_repo_agents.py. - New shell/Makefile entry points to run inspection agents from a single command.
- Continual-learning Cursor skill +
agents-memory-updatersubagent, with synced docs (AGENTS.md, scripts/README.md, REPO_AUTOMATION_GUIDE.md).
Reviewed changes
Copilot reviewed 10 out of 10 changed files in this pull request and generated 2 comments.
Show a summary per file
| File | Description |
|---|---|
scripts/agents/agents_md_audit_agent.py |
New audit agent; follows the existing AutomationAgent/CLI conventions for parsing, findings, status, and write_status. |
tests/test_agents_md_audit_agent.py |
New tests covering valid/missing/over-limit/secret/stale/CLI paths; fixture embeds a wall-clock-relative date that will eventually flip the "valid" case to a warning. |
scripts/run_repo_agents.py |
Registers the new agent module so it participates in standard runs. |
scripts/run_ai_automation.sh |
New wrapper that hardcodes .venv/bin/python without the venv-fallback used by sibling scripts. |
Makefile |
Adds agents, agents-dry, ai-automation targets and .PHONY entries (help preserved). |
scripts/README.md |
Documents the new agent, entry points, and corrected --run-agents ordering. |
docs/guides/REPO_AUTOMATION_GUIDE.md |
Documents the new agent and corrected (after-gate) --run-agents ordering, verified against code. |
AGENTS.md |
Adds repo-agent and continual-learning usage notes to test/entry-point sections. |
.cursor/skills/continual-learning/SKILL.md |
Orchestration-only skill delegating to the updater subagent. |
.cursor/agents/agents-memory-updater.md |
Subagent definition for transcript-mining and AGENTS.md memory updates. |
|
|
||
| ## Learned Workspace Facts | ||
|
|
||
| - Unit test suite passes 2700 tests (as of 2026-06-20). |
Comment on lines
+3
to
+4
| REPO_ROOT="$(cd "$(dirname "$0")/.." && pwd)" | ||
| exec "${REPO_ROOT}/.venv/bin/python" "${REPO_ROOT}/scripts/run_repo_agents.py" "$@" |
Bugbot couldn't run - usage limit reachedBugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit. A user or team admin can review and increase usage limits in the Cursor dashboard. (requestId: serverGenReqId_6abfa7a0-2ede-4e0a-bdbe-6879a754ee3b) |
|
You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard. |
![sourcery-ai[bot]](https://avatars.githubusercontent.com/in/48477?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Hey - I've left some high level feedback:
- The
test_valid_agents_md_passesfixture hardcodes a2026-06-20date and uses the defaultstale_date_days=30, so as time passes this will start failing due to the stale-date check; consider deriving the date dynamically (e.g., usingdatetime.now()minus a small offset) or overridingstale_date_daysto a very large value in that test to keep it stable over time.
Prompt for AI Agents
Please address the comments from this code review:
## Overall Comments
- The `test_valid_agents_md_passes` fixture hardcodes a `2026-06-20` date and uses the default `stale_date_days=30`, so as time passes this will start failing due to the stale-date check; consider deriving the date dynamically (e.g., using `datetime.now()` minus a small offset) or overriding `stale_date_days` to a very large value in that test to keep it stable over time.Help me be more useful! Please click 👍 or 👎 on each comment and I'll use the feedback to improve your reviews.
🔐 CodeQL — Open Alerts on this PR
Copilot Autofix suggestions (if enabled) appear as inline review comments on the affected lines. |
Bugbot couldn't run - usage limit reachedBugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit. A user or team admin can review and increase usage limits in the Cursor dashboard. (requestId: serverGenReqId_e3ecee1a-e640-4045-b89d-428a061e630b) |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Adds the continual-learning Cursor workflow (skill + memory-updater subagent), a new read-only
agents-md-auditinspection agent for AGENTS.md Learned sections, shell/Makefile shortcuts to run repo agents, and documentation fixes for the automation stack.Changes
.cursor/agents/agents-memory-updater.mdand.cursor/skills/continual-learning/SKILL.mdfor transcript mining and AGENTS.md memory updatesscripts/agents/agents_md_audit_agent.py(agents-md-audit) — validates Learned section structure, bullet limits, secret patterns, and stale datesscripts/run_repo_agents.pyscripts/run_ai_automation.shand Makefile targets (make agents,make agents-dry,make ai-automation)--run-agentsstep order to match code)~/.cursor/projects/workspace/canvases/automation-status.canvas.tsx(IDE-local artifact)Verification
pytest tests/test_agents_md_audit_agent.py -v— 9 passed./scripts/run_ai_automation.sh --agent agents-md-audit --dry-run— okmake agents-dry— 4 agents run (2 ok, 2 warning from pre-existing marker/docstring findings)/continual-learningin chat to delegate toagents-memory-updaterSummary by Sourcery
Introduce a new automation agent for auditing AGENTS.md and wire it into the repo automation workflow, alongside continual-learning support and new entry points for running agents.
New Features:
Enhancements:
Tests: