feat: Claude Code extended features — Issue Implementation, Doc Sync, Security Review, Checklists, Maintenance

[Kavirubc]

Summary

Extend the Claude Code @simili-bot integration (from #105 / PR #106) with 5 additional features, all configurable via .github/simili.yaml.

Features

1. Issue Implementation (label-triggered)

When a maintainer adds the implement label to an issue, Claude Code reads the issue description and creates a PR with the implementation. Only org OWNER/MEMBER/COLLABORATOR can trigger.

2. Doc Sync

When a PR changes files matching configured paths (e.g., src/api/**), Claude auto-updates docs to match.

3. Security Review (selective)

OWASP-aligned security analysis on PRs — only triggered via label (security-review) to avoid wasting tokens on every PR.

4. Custom Review Checklist

Users define a review checklist in simili.yaml. When triggered (via label or @simili-bot review), Claude checks the PR against the checklist.

5. Scheduled Maintenance

Weekly/monthly automated tasks: dependency audits, stale issue cleanup, TODO scanning. Runs on cron schedule.

Configuration

All features live under claude_code: in .github/simili.yaml:

claude_code:
  enabled: true
  trigger_phrase: "@simili-bot"

  issue_implement:
    enabled: true
    trigger_label: "implement"

  doc_sync:
    enabled: true
    watch_paths: ["src/api/**"]
    doc_paths: ["docs/"]

  security_review:
    enabled: true
    trigger_label: "security-review"

  review_checklist:
    enabled: true
    items:
      - "No hardcoded credentials"
      - "Unit tests for new functions"

  maintenance:
    enabled: true
    tasks:
      - "Check for outdated dependencies"
      - "Scan for TODO/FIXME comments"

Implementation Approach

Each feature follows the same pattern as the existing @simili-bot handler:

1. Go code detects the trigger condition
2. Validates authorization
3. Builds the prompt with context
4. Writes GitHub Actions outputs
5. Conditional workflow step invokes claude-code-action

Depends on: #105 (PR #106)

[gh-simili-bot]

Simili Triage Report

Note

Quality Score: 8.5/10 (Good)
The issue could be improved. See suggestions below.

Classification

Category	Value
Labels

Quality Improvements

• Some feature descriptions lack specific details on expected output or exact behavior (e.g., Doc Sync, Security Review, Custom Review Checklist, Scheduled Maintenance)
• The last sentence of the 'Scheduled Maintenance' feature description is incomplete
• Provide more specific details for each feature's expected output or behavior (e.g., for Doc Sync, clarify how docs are 'updated to match'; for Security Review, specify the analysis output; for Custom Review Checklist, explain how Claude 'checks' the PR; for Scheduled Maintenance, detail the actions for each task)
• Complete the description for the 'Scheduled Maintenance' cron schedule

Similar Threads

Similarity	Type	Thread	Status
85%	🔀	#106 feat: integrate Claude Code @simili-bot for AI-...	Open
85%	📝	#105 feat: Add Claude Code GitHub Action for AI-assi...	Open
81%	🔀	#1 feat: Initial implementation of Simili-bot v0.0.1	—

Note

Related Issues Found
This issue is not a duplicate, but is related to #106 (feat: integrate Claude Code @simili-bot for AI-assisted PR fixes), #105 (feat: Add Claude Code GitHub Action for AI-assisted PR fixes via @simili-bot comments), #1 (feat: Initial implementation of Simili-bot v0.0.1).

---

_{Generated by Simili Bot}