deslop
Pass
Audited by Gen Agent Trust Hub on Jul 9, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists entirely of natural language instructions for the agent to identify and remove unnecessary comments, defensive checks, and other code patterns deemed 'slop'. It does not contain any executable scripts, configuration for network access, or file system manipulation outside of standard agent operations.
- [INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it processes branch diffs (untrusted data).
- Ingestion points: Branch diffs (referenced in SKILL.md).
- Boundary markers: Absent.
- Capability inventory: The skill facilitates code modification based on the processed diffs.
- Sanitization: Absent.
- Note: The guardrails requiring the agent to 'Keep behavior unchanged' and 'Prefer minimal, focused edits' act as a mitigation against adversarial content within the diffs.
Audit Metadata