debugging-codeceptjs-tests

Pass

Audited by Gen Agent Trust Hub on Aug 25, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill describes a workflow that involves reading and analyzing external files such as HTML snapshots and test traces. These artifacts originate from the web pages being tested and could contain adversarial instructions intended to influence the agent's behavior.
  • Ingestion points: The agent reads output/trace.md, output/console.json, and HTML snapshots (*.html) via tools like codeceptq and snapshot.
  • Boundary markers: The skill suggests using codeceptq to limit the amount of HTML loaded into the context, but no explicit security boundary markers or instructions to ignore embedded commands are present.
  • Capability inventory: The agent has access to run_code (arbitrary JavaScript execution in the browser), run_test, and snapshot tools.
  • Sanitization: No explicit sanitization of the HTML or trace content is mentioned before it is processed by the agent.
  • [DYNAMIC_EXECUTION]: The skill workflow relies on the run_code tool to execute arbitrary JavaScript snippets within the live browser session for troubleshooting. While this is an intended diagnostic feature, it represents a high-capability tool that requires careful use and user oversight.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 25, 2026, 07:32 AM
Security Audit — agent-trust-hub — debugging-codeceptjs-tests