refactor-codebase
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by instructing the agent to ingest and act upon data from an external codebase that may contain malicious instructions.
- Ingestion points: The agent reads the codebase structure and individual file contents to identify module boundaries, coupling hotspots, and to update imports (seen in
SKILL.mdworkflow steps 1 and 4). - Boundary markers: There are no instructions to use delimiters or specific warnings to ignore instructions embedded within the source code being refactored.
- Capability inventory: The skill guides the agent to run verification commands such as
tsc,npm run build,pytest, andcargo check, or to perform direct shell operations likesedandnode -e(REFERENCE.mdVerification Commands section). - Sanitization: The instructions do not include steps to sanitize the codebase content or validate the integrity of the project configuration files (e.g.,
package.jsonscripts) before executing the suggested verification commands.
Audit Metadata