The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes several shell scripts designed to automate developer workflows through command execution.
Evidence: auto-pr-create.sh, issue-triage.sh, release-automation.sh, and workflow-monitor.sh execute gh, git, and build/test tools including npm, make, and pytest.
[DATA_EXFILTRATION]: The skill provides guidance on managing local files associated with GitHub authentication and security.
Evidence: references/advanced-features.md includes instructions for adding SSH public keys from ~/.ssh/id_ed25519.pub to a GitHub account.
Evidence: references/troubleshooting.md demonstrates authenticating with the CLI using tokens stored in a local token.txt file.
[PROMPT_INJECTION]: An automated triage script processes untrusted external data, creating a surface for indirect prompt injection.
Ingestion points: Issue titles and bodies are fetched from the GitHub API in examples/issue-triage.sh.
Boundary markers: None; the script processes fetched content directly using grep and jq.
Capability inventory: The script performs resource management actions including labeling (gh issue edit --add-label) and assigning issues (gh issue edit --add-assignee).
Sanitization: Content is parsed as structured JSON using jq prior to keyword matching, which provides basic data integrity but does not sanitize against instructional content.

cli-gh