audit
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection because it ingests untrusted data from the codebase to drive automated refactoring actions.
- Ingestion points: The skill reads project metadata files (
package.json,go.mod,Gemfile,pyproject.toml,pom.xml,Cargo.toml), project convention files (CLAUDE.md,AGENTS.md), decision logs (team-memory/MEMORY.md), and the source code itself across the entire codebase. - Boundary markers: Absent. The instructions do not provide delimiters or warnings to ignore instructions that might be embedded in the code or documentation being audited.
- Capability inventory: The skill possesses significant capabilities including the ability to delete files, modify source code, and execute version control commands (
git add,git commit). - Sanitization: Absent. The skill does not describe any sanitization, validation, or filtering of the content read from the files before processing it.
- [COMMAND_EXECUTION]: The skill executes shell commands to determine the scope of changes and persist refactoring results.
- Evidence: Usage of
git diff,git add -A, andgit commitin Step 1 and Step 7.
Audit Metadata