explain-yesterday-changes
Pass
Audited by Gen Agent Trust Hub on Jul 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local git commands to analyze the repository state. Evidence:
git rev-parse --git-dir,git log --stat, andgit log --patchare used inSKILL.mdto retrieve history and diffs.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from the local git repository which could contain adversarial instructions embedded in commit messages or code diffs.\n - Ingestion points:
git logoutput containing commit messages and file diffs inSKILL.md.\n - Boundary markers: Absent. There are no delimiters or specific instructions to the agent to disregard potential instructions within the git output.\n
- Capability inventory: The skill performs local read-only git operations and generates text summaries; it lacks tools for network exfiltration or file writing in the provided context.\n
- Sanitization: Absent. The skill does not specify any filtering or validation of the content retrieved from git history.
Audit Metadata