- name
- git-log-intelligence
- description
- Fetch, filter, and summarize GitHub repository activity without cloning. Use whenever the user asks what changed in a repo, wants a changelog summary, asks about recent commits or PRs, or wants to tune out noisy commit patterns. Maintains a persistent ignore list to hide bot/CI/chore noise automatically.
- metadata
- { "openclaw": { "requires": { "bins": ["gh"], "env": ["GITHUB_PERSONAL_ACCESS_TOKEN"] } } }
Prerequisites
ghCLI installed. Authenticate via one of:
- GITHUB_PERSONAL_ACCESS_TOKEN environment variable (required)
- For read-only public repo access, grant only the
public_reposcope.
For private repos, use repo scope.
- Note: this skill creates
.config/git_filters.jsonrelative to the script's
location to persist ignore patterns. Delete this file to reset to defaults.
Compatibility
- Required tools:
gh - Environment variables:
- GITHUB_PERSONAL_ACCESS_TOKEN (required): GitHub Personal Access Token. Grant minimal scopes: public_repo for public repos, repo for private repos.
Directory Structure
git_log_intelligence.py: The primary script containing execution logic..config/git_filters.json: The persistent memory of ignore patterns.
Capabilities
1. Summarize Repo Changes
Fetches the last N days of commits and filters them against the "Ignore List." Command: python3 skills/git_log_intelligence.py summarize OWNER/REPO DAYS
Agent Logic:
Call the script with the repo name and timeframe. Receive a filtered list of "Important" commits. Present a natural language summary to the user, noting how many noisy commits were hidden.
2. Maintaining and Updating the Ignore List
The agent can add new patterns to the ignore list based on user feedback. Command: python3 skills/git_log_intelligence.py ignore "PATTERN"
The agent can also show the current ignore patterns or remove specific ones. Command: python3 skills/git_log_intelligence.py show
The agent can remove a pattern with: Command: python3 skills/git_log_intelligence.py remove "PATTERN"
Usage Example:
1. User: "Claw, stop showing me commits from 'GitHub Actions' or anything starting with 'ci:'." 2. Agent: Runs python3 skills/git_log_intelligence.py ignore "GitHub Actions" and python3 skills/git_log_intelligence.py ignore "^ci:".
3. User: "Show me the current ignore patterns." 4. Agent: Runs python3 skills/git_log_intelligence.py show
5. User: "Remove the ignore pattern for 'GitHub Actions'." 6. Agent: Runs python3 skills/git_log_intelligence.py remove "GitHub Actions"
Technical implementation:
See the git_log_intelligence.py script for the full code, but the key functions are:
- Quick Scan (Default):
python3 skills/git_log_intelligence.py summarize OWNER/REPO DAYS
- Deep Context (Full Messages):
python3 skills/git_log_intelligence.py summarize OWNER/REPO DAYS --full
Note to Agent:
Use --full when the user asks for a "detailed summary," "technical breakdown," or "why" things were changed. Use the default mode for a "quick list" or "recent activity check." You can also ask the user if they want the full context after showing the quick summary.
How It Works
The script uses gh api to fetch commits via the GitHub REST API (/repos/{owner}/{repo}/commits), filters them client-side against .config/git_filters.json, and outputs structured text for the agent to summarize.
Troubleshooting
If the script exits non-zero, check gh auth status first.
User Interaction Flow
1. Request: "What's new in tailscale/tailscale this week?" 2. Execution: Agent runs python3 skills/git_log_intelligence.py summarize tailscale/tailscale 7. 3. Feedback: "Too much noise from the build bot." 4. Learning: Agent runs python3 skills/git_log_intelligence.py ignore "build bot". 5. Next Time: The build bot commits are gone forever.