codebase-litter-audit
Pass
Audited by Gen Agent Trust Hub on Jun 25, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a bundled Python script (
scan_litter.py) and standard command-line tools such asgitandghto perform the audit and check repository status. - [PROMPT_INJECTION]: The skill audits untrusted codebase data, which creates a surface for indirect prompt injection. Malicious instructions hidden in comments or strings within the audited repository could attempt to influence the agent's behavior during the verification step.
- Ingestion points: The
scan_litter.pyscript reads the content of all text files in the target repository, and the agent manually reads files identified as findings. - Boundary markers: Absent; instructions do not specify the use of delimiters or 'ignore' instructions to isolate audited code from the agent's operational context.
- Capability inventory: Shell access (
bash) and file system read/write permissions. - Sanitization: Absent; the scanner performs regex-based identification without sanitizing content for LLM ingestion.
Audit Metadata