brainstorming

Pass

Audited by Gen Agent Trust Hub on Mar 27, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill launches a local Node.js server using shell scripts (start-server.sh, stop-server.sh) to render and serve visual design elements in a browser.\n- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests project files and documentation to understand the existing context, which could contain malicious instructions.\n
  • Ingestion points: Project files, documentation, and commit history are read from the file system during the initial exploration phase defined in SKILL.md.\n
  • Boundary markers: There are no explicit markers or delimiters used to separate project data from the agent's primary instructions.\n
  • Capability inventory: The skill can write files to the project directory, run a local web server, and trigger subagent reviews.\n
  • Sanitization: No validation or sanitization of the ingested project data is performed before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 27, 2026, 01:11 PM