The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the GitHub CLI (gh) to perform several operations, including checking PR status and retrieving logs from external CI runs.
Evidence: gh pr checks {{PR_NUMBER}} --json ... and gh run view {{RUN_ID}} --job {{JOB_ID}} --log in SKILL.md.
[COMMAND_EXECUTION]: The skill provides procedures for system-level tool installation and authentication.
Evidence: brew install gh and gh auth login commands listed in references/gh-installation.md.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it processes untrusted content from external sources (logs and diffs) which could contain adversarial instructions.
Ingestion points: CI logs retrieved via gh run view and PR diffs accessed through {{PR_URL}}.diff.
Boundary markers: Absent; there are no specific delimiters or instructions for the agent to ignore embedded commands within the logs or code diffs.
Capability inventory: The skill has the ability to execute shell commands and read external network data via the gh tool.
Sanitization: No sanitization or filtering logic is present to ensure that data from logs or diffs does not influence agent behavior beyond the intended analysis.
[EXTERNAL_DOWNLOADS]: The skill suggests downloading the GitHub CLI using Homebrew, a well-known and trusted package management service.
Evidence: brew install gh in references/gh-installation.md.

review-ci