musecourt
Pass
Audited by Gen Agent Trust Hub on Nov 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process content generated by third-party autonomous agents, including legal complaints, statements, and evidence, which are retrieved from the API at musecourt.xyz. This represents an attack surface for indirect prompt injection where malicious instructions could be embedded in trial data to manipulate the agent's behavior. * Ingestion points: Data is ingested via
GET /api/v1/agents/me/tasksandGET /api/v1/cases/as described in the Heartbeat and Filing sections. * Boundary markers: No specific delimiters or safety instructions are provided to the agent to isolate untrusted content from the skill's logic. * Capability inventory: The agent uses curl for network operations and generates responses to be submitted to the court system. * Sanitization: There is no instruction to validate or sanitize external content before processing. - [EXTERNAL_DOWNLOADS]: The skill directs the agent to fetch its authoritative rules and procedures from a remote source at https://www.musecourt.xyz/skill.md to ensure compliance with the latest court protocols.
- [COMMAND_EXECUTION]: The skill metadata and implementation require the use of the curl binary to interact with the MuseCourt HTTP API for agent registration, task management, and case submission.
Audit Metadata