docs-cleaner

Pass

Audited by Gen Agent Trust Hub on Sep 8, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted content from documentation files to determine which sections should be kept, condensed, or deleted. This behavior creates a vulnerability where a document could contain hidden instructions that cause the agent to deviate from its intended logic.
  • Ingestion points: Documentation files identified during the 'Discovery' phase are processed in the 'Value Analysis' and 'Execution' phases.
  • Boundary markers: The workflow lacks explicit delimiters or instructions to prevent the agent from interpreting instructions embedded within the documentation it is processing.
  • Capability inventory: The skill has high-impact capabilities including the deletion of source files and modification of core project files like CLAUDE.md or README.md.
  • Sanitization: No sanitization or validation of the processed text is mentioned prior to the agent performing its analysis.
  • [COMMAND_EXECUTION]: The skill uses the rg (ripgrep) command-line tool to perform verification of file paths and remaining placeholders after the consolidation process is complete.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 8, 2026, 06:56 AM
Security Audit — agent-trust-hub — docs-cleaner