code-overhaul-review
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEMETADATA_POISONINGINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [METADATA_POISONING]: The SKILL.md file identifies as version 1.0.0, but the included CHANGELOG.md describes version 2.1.0 as the current state following a full rewrite. As a result, the skill lacks numerous safety protocols, environment preflight checks, and fallback mechanisms defined in the changelog, which could lead to inconsistent agent behavior.
- [INDIRECT_PROMPT_INJECTION]: The skill audits untrusted local code repositories, creating an attack surface for indirect prompt injection. • Ingestion points: Auditing logic reads repository files via the Read, Grep, and Glob tools. • Boundary markers: There are no instructions for the agent to isolate or ignore instructions within the code files being analyzed. • Capability inventory: Access to the Bash tool to execute audit commands and the bd task management tool. • Sanitization: There is no evidence of data sanitization before findings from the codebase are processed into reports or commands.
- [COMMAND_EXECUTION]: The skill utilizes the Bash tool to run standard development utilities (e.g., git, swiftlint, go vet) and a custom task management command named bd.
Audit Metadata