tend-refactor
Pass
Audited by Gen Agent Trust Hub on Aug 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted source code from the repository, creating an attack surface for indirect prompt injection.
- Ingestion points: The skill reads files from the project's src/, lib/, routes/, and components/ directories.
- Boundary markers: The instructions define off-limits paths (e.g., .env, CI configs) but do not specify the use of delimiters or instructions to ignore embedded commands in the source code.
- Capability inventory: The skill can execute shell commands (lint, typecheck, tests), run a local validation script, and write refactored code to the disk.
- Sanitization: There is no evidence of sanitization or filtering of the ingested content.
- [COMMAND_EXECUTION]: The skill performs automated execution of local project tools and internal scripts.
- Evidence: Phase 3 and Phase 4 of the instructions require running 'typecheck', 'lint', and project-specific tests.
- Evidence: The skill runs the local script 'node scripts/validate-finding.mjs' to verify that findings meet the required criteria.
Audit Metadata