skills/liangmiqwq/skills/kill-ai-slop/Gen Agent Trust Hub

kill-ai-slop

Pass

Audited by Gen Agent Trust Hub on Aug 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is a self-contained code analysis utility that operates locally within the project environment. It does not utilize external dependencies or perform network communication, significantly reducing the risk of data exfiltration or remote code execution.
  • [PROMPT_INJECTION]: The skill implements a workflow for processing untrusted local source files to detect 'AI slop' patterns. 1. Ingestion points: scripts/scan.mjs reads file content. 2. Boundary markers: The tool supports deslop-ignore comments for suppression. 3. Capability inventory: The agent is instructed to modify files based on the scan results. 4. Sanitization: Not applicable as the data is used for regex matching rather than execution. This ingestion surface is managed through human-in-the-loop verification steps in the workflow.
  • [SAFE]: The scripts/scan.mjs script utilizes a dynamic import mechanism to load optional rule sets via the --rules flag. This is a standard extension feature for developer tools and is restricted to local files provided by the user or agent during execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 6, 2026, 02:48 PM
Security Audit — agent-trust-hub — kill-ai-slop