legal-research

Warn

Audited by Gen Agent Trust Hub on Jul 17, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands to modify the local environment and install tools. Specifically, it directs the use of pip install and claude mcp add to set up required MCP servers.
  • [REMOTE_CODE_EXECUTION]: The skill utilizes a remote MCP server via a URL (https://tlr.dr-lawbot.com/mcp). This configuration allows the agent to call tools hosted on an external server, which governs how the agent processes legal queries.
  • [EXTERNAL_DOWNLOADS]: The 'Bootstrapping' section triggers the download and installation of external Python packages from PyPI (mcp-taiwan-legal-db) and the registration of remote services.
  • [DATA_EXFILTRATION]: Legal queries and case details are transmitted to an external domain (tlr.dr-lawbot.com) as part of the tool's semantic search functionality. This involves sending user-provided context to a third-party infrastructure.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 17, 2026, 08:43 PM
Security Audit — agent-trust-hub — legal-research