docs-cleaner
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides several shell commands for documentation discovery and management, including git, ripgrep (rg), and wc.
- [DYNAMIC_EXECUTION]: The skill includes a Python script string intended for execution via python3 -c to verify that content was preserved during documentation consolidation.
- Evidence: SKILL.md contains a Python one-liner using pathlib and sys to count occurrences of specific strings in documentation files.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted documentation files within a repository, creating a surface for indirect prompt injection if those files contain malicious instructions.
- Ingestion points: Markdown and text files located within the repository search boundary.
- Boundary markers: The skill does not implement explicit delimiters or warnings to separate processed documentation content from agent instructions.
- Capability inventory: File system access, including moving (git mv), writing, and deleting documentation files.
- Sanitization: No content filtering or sanitization is performed on the ingested documentation files.
- [DATA_EXFILTRATION]: The skill suggests using the lychee tool, which can perform outbound network requests to verify external URLs found in documentation.
Audit Metadata