pr-changelog
Pass
Audited by Gen Agent Trust Hub on Mar 14, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it reads and processes raw data from git diffs. If the code changes being summarized contain malicious instructions (e.g., within code comments), the agent could be manipulated into producing misleading changelog entries or bypassing intended behavior.
- Ingestion points: Output of
git diff main...HEADin Step 1. - Boundary markers: None. There are no instructions to the agent to treat the diff output as untrusted data or to ignore embedded commands.
- Capability inventory: Execution of
git diffto read repository state. - Sanitization: None. The skill directly processes the diff output.
- [COMMAND_EXECUTION]: The skill executes
git diff main...HEADto retrieve the data necessary for generating a changelog. This is a standard and expected operation given the skill's stated purpose.
Audit Metadata