ship-it
Pass
Audited by Gen Agent Trust Hub on Sep 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read repository-specific documentation and CI configurations to determine which verification commands to execute.
- Ingestion points: CI configuration files and contributor documentation (SKILL.md, Step 3).
- Boundary markers: Absent; the agent is directed to follow instructions found in the repo without specific delimiters.
- Capability inventory: Shell command execution for builds/tests, git operations, and GitHub CLI access (SKILL.md, Steps 1-5).
- Sanitization: Absent; the agent is prompted to inline transcripts directly.
- [DATA_EXFILTRATION]: The workflow requires the agent to capture and upload command transcripts, logs, and screenshots as "testing proof" into the body of pull requests. This could lead to the unintended exposure of environment variables, local file paths, or credentials if they appear in command outputs.
- [COMMAND_EXECUTION]: The agent is authorized to execute arbitrary commands locally that are derived from the repository's own build and test logic. This behavior relies on the integrity of the local environment and the repository's configuration files.
Audit Metadata