governance-autoresearchlisted

# Governance Autoresearch Skill ## Overview Adapts Karpathy's autoresearch pattern to governance file improvement: - **Research** each file's topics on X (last 7 days) - **Cross-reference** with bookmarks data (if available in `X-Bookmarks Import/`) - **Propose** ONE atomic change per file - **Validate** structure and SSOT consistency - **Keep** if it improves the file, **discard** if not - **Commit** each kept change with `experiment:` prefix ## Prerequisites - `X_BEARER_TOKEN` in `.env` - Python 3.10+ - Bookmarks import data (optional, in `X-Bookmarks Import/`) ## Workflow — Execute These Steps Per File ### Step 0: Get Research Context Run the research script for the target file: ```bash python3 "X-Bookmarks Import/skills/governance-autoresearch/scripts/governance_research.py" "<file_path>" ``` This outputs JSON with: - Top X posts about the file's topics (engagement-ranked) - External links being shared - File summary for context ### Step 1: Read the File Read the full governance file. Note: - Current structure and sections - YAML frontmatter (doc_type, ssot_owner) - Key policies, rules, or patterns defined ### Step 2: Cross-Reference Bookmarks (if available) Check `X-Bookmarks Import/data/articles-batch-1.md`, `data/articles-batch-2.md`, `data/github-repos.md`, `data/other-sources.md` for content relevant to this file's domain. Key sources: | Bookmark Content | Relevant To | |-----------------|-------------| | Claude Code Skills (Anthropic) | skill-standards