changelog
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes commit messages retrieved from the git history, which are considered untrusted external data. This creates a potential surface for indirect prompt injection if a commit message contains malicious instructions intended to subvert the agent's behavior during the changelog creation process.
- Ingestion points: The skill executes
git log main..HEAD --onelineto ingest commit messages (SKILL.md). - Boundary markers: There are no instructions or delimiters defined to ensure the agent ignores or isolates embedded instructions within the processed commit data.
- Capability inventory: The skill writes new files to the
changelog/directory and invokes the/prose-reviewtool (SKILL.md). - Sanitization: No sanitization, escaping, or filtering of the external commit content is performed before processing.
- [COMMAND_EXECUTION]: The skill executes local git commands to analyze the repository state.
- Evidence: Implementation of
git log main..HEAD --oneline(SKILL.md).
Audit Metadata