skills/nmnmcc/skills/point/Gen Agent Trust Hub

point

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The workflow incorporates the execution of software checks to verify changes.
  • Evidence: Step 5 of the 'Run the loop' section directs the agent to 'Run the most direct check and the nearest useful regression check.'
  • [INDIRECT_PROMPT_INJECTION]: The skill analyzes and modifies source code, which is a surface for indirect prompt injection.
  • Ingestion points: Reads software source code and task context in SKILL.md.
  • Boundary markers: No specific delimiters or instructions to ignore embedded commands are present.
  • Capability inventory: Includes capabilities for file modification and command execution for testing.
  • Sanitization: No explicit validation or filtering of the ingested code content is described.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 01:07 AM
Security Audit — agent-trust-hub — point