babysit
Pass
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
gh(GitHub CLI) tool to perform operations such as viewing PR details (gh pr view) and monitoring CI status (gh pr checks --watch). - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and act upon untrusted external data from pull request comments and CI failure logs.
- Ingestion points: The agent reads PR comments and logs from external reviewers (bots or humans) in
SKILL.mdto triage actions. - Boundary markers: There are no technical delimiters or instructions to treat external data as untrusted text strings to prevent the agent from following instructions embedded within comments.
- Capability inventory: The skill possesses the capability to modify the local filesystem (applying fixes) and execute network-enabled commands via
gh(pushing code, replying to comments). - Sanitization: While the skill instructs the agent to treat comments as "arguments, not instructions," there is no automated sanitization or filtering of the input data.
Audit Metadata