skills/andrewgleave/skills/cleanser/Gen Agent Trust Hub

cleanser

Pass

Audited by Gen Agent Trust Hub on Mar 18, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute build and lint commands found within the target project environment (run the project's lint or build command if one exists). If the project contains malicious configuration files (e.g., package.json, Makefile), this could result in the execution of unintended commands.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection, where malicious instructions hidden in the code being reviewed could influence the agent's behavior.
  • Ingestion points: Source code files and project directories provided by the user for review.
  • Boundary markers: Absent. The sub-agent receives the target code and instructions without explicit delimiters to separate data from commands.
  • Capability inventory: The agent can perform file write operations (Make all changes directly) and execute system commands (run the project's lint or build command).
  • Sanitization: Absent. The skill does not implement validation or filtering for the content of the files being processed.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 18, 2026, 12:02 PM