anti-slop-fix

Pass

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes arbitrary source code files provided via user input.
  • Ingestion points: Source code content read during the audit workflow (SKILL.md, Step 1).
  • Boundary markers: The instructions do not define delimiters or explicit warnings for the agent to ignore instructions embedded within the files being processed.
  • Capability inventory: The skill utilizes an 'Edit tool' to apply changes to the file system (SKILL.md, Step 3 and 4).
  • Sanitization: No sanitization or validation of the file content is performed to filter out potential malicious instructions before processing.
  • [COMMAND_EXECUTION]: The skill output includes instructions for the user to execute local shell commands (npm run lint && npm test) to verify changes. While standard for development, this involves executing code in the local environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 13, 2026, 06:34 AM