skills/udecode/plate/hard-cut/Gen Agent Trust Hub

hard-cut

Pass

Audited by Gen Agent Trust Hub on Apr 4, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill contains instructional guidance for feature removal without any attempts to override safety protocols or system constraints.
  • [PROMPT_INJECTION]: The skill processes user requests for code deletion, representing a functional surface for indirect prompt injection. 1. Ingestion points: User-provided prompts such as 'hard cut' or 'rip it out' (SKILL.md). 2. Boundary markers: None specified. 3. Capability inventory: File system modification and deletion (implicit). 4. Sanitization: None specified.
  • [COMMAND_EXECUTION]: The skill references the use of 'rg' (ripgrep) to trace code references, which is a standard development tool used here for legitimate codebase searching.
  • [DATA_EXFILTRATION]: No network access, credential harvesting, or sensitive data exposure patterns were identified.
  • [REMOTE_CODE_EXECUTION]: The skill does not perform any external package installations or execute remote scripts.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 4, 2026, 01:31 AM