ai-slop-cleaner
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill defines a defensive workflow focused on code simplification and 'anti-slop' cleanup. It explicitly prioritizes regression testing and behavior locking before any code modifications occur.
- [INDIRECT_PROMPT_INJECTION]: The skill acts on external codebases which represents a theoretical attack surface. However, the risk is mitigated by explicit instructions for verification and reviewer passes.
- Ingestion points: Target files and directories provided as arguments to the cleaner (SKILL.md).
- Boundary markers: Instructions mandate 'Preserve behavior unless the user explicitly asks for behavior changes' and 'Lock behavior with focused regression tests first'.
- Capability inventory: File system inspection, modification (edits/deletions), and execution of local quality gates (lint, typecheck, tests).
- Sanitization: Multi-pass verification, reviewer-only mode (
--review), and evidence-dense reporting requirements serve as procedural sanitization.
Audit Metadata