brainstorming
Pass
Audited by Gen Agent Trust Hub on Nov 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill manages a local development server using shell scripts and Node.js.
- Evidence: The scripts
scripts/start-server.shandscripts/server.cjsare used to launch a background process on the user's machine to serve visual brainstorming aids. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from local browser interactions which could theoretically be tampered with, though the risk is mitigated by localhost binding.
- Ingestion points: The agent reads user interaction data (clicks/selections) from a
.eventsfile located in the session directory (e.g.,scripts/server.cjs). - Boundary markers: None explicitly defined in the file-reading logic; the agent parses the JSON lines directly.
- Capability inventory: The skill has the ability to start/stop a local HTTP server and perform file read/write operations within the designated session directory.
- Sanitization: The Node.js server implementation uses
path.basenamewhen serving files to prevent directory traversal attacks, and the server binds to127.0.0.1by default to prevent external access.
Audit Metadata