skills/iml1s/twmcp/tw-utils/Gen Agent Trust Hub

tw-utils

Pass

Audited by Gen Agent Trust Hub on Jun 18, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill recommends installing the twmcp package from the author's GitHub repository (iml1s/twmcp) or via pipx. These are vendor-owned resources directly associated with the skill's purpose.
  • [COMMAND_EXECUTION]: Provides setup instructions that involve executing shell commands to install the necessary MCP server and add it to the agent configuration (e.g., pipx install twmcp, claude mcp add).
  • [PROMPT_INJECTION]: The skill processes untrusted user data, creating a surface for indirect prompt injection. * Ingestion points: User-provided identity numbers (IDs), tax IDs, and address strings defined in SKILL.md. * Boundary markers: Absent; there are no instructions for the agent to use delimiters or ignore embedded commands when passing data to the tools. * Capability inventory: Leverages external MCP tools for validation, generation, and normalization (e.g., validate_taiwan_id_number, normalize_taiwan_address). * Sanitization: Absent; the skill does not specify any validation or filtering of user input before it is sent to the twmcp server.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 18, 2026, 05:35 PM
Security Audit — agent-trust-hub — tw-utils