skills/first-fluke/mimic/qa-agent/Gen Agent Trust Hub

qa-agent

Pass

Audited by Gen Agent Trust Hub on Apr 2, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill defines a professional and structured QA workflow using established security protocols and audit tools like npm audit, bandit, and lighthouse.\n- [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection because it is designed to analyze untrusted external codebase content.\n
  • Ingestion points: Codebase files and modules identified through scoping tools such as search_for_pattern and get_symbols_overview in resources/execution-protocol.md.\n
  • Boundary markers: Absent. The execution protocol does not specify delimiters or instructions to isolate untrusted code content from the agent's control logic.\n
  • Capability inventory: The skill uses tool calls for reading files, searching patterns, and executing CLI-based auditing utilities.\n
  • Sanitization: Absent. External codebase content is reviewed directly without prior sanitization or instruction-filtering.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 2, 2026, 07:25 PM
Security Audit — agent-trust-hub — qa-agent