skills/cursor/plugins/X Money guide/Gen Agent Trust Hub

X Money guide

Pass

Audited by Gen Agent Trust Hub on Oct 22, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes and relays message strings returned by the X Money MCP server, creating a potential surface for instructions embedded in external data to influence agent behavior.
  • Ingestion points: The agent reads the message field from tool outputs provided by the X Money MCP server (https://mcp.money.x.com/mcp).
  • Boundary markers: No specific technical delimiters are defined for the tool output, although the skill provides natural language instructions on how to interpret and relay these messages.
  • Capability inventory: The agent has the capability to perform financial operations including sending money, requesting money, and card creation as defined by the MCP server.
  • Sanitization: The skill includes explicit prohibitions against printing or storing sensitive data (e.g., card numbers, CVC, expiry) and instructs the agent to use its own voice rather than quoting technical internals, which serves as a manual sanitization step.
  • [PROMPT_INJECTION]: The skill contains instructions to suppress internal technical details (e.g., "Relay it in your own words instead of describing internals" and "Do not explain OAuth, tokens... or how the connection works internally"). While these instructions involve concealment of information from the user, they are identified as benign user experience and security guidelines intended to prevent technical jargon and potential credential metadata from reaching the end user.
  • [SAFE]: The skill is primarily composed of defensive security instructions designed to protect the user's financial assets and credentials. It explicitly forbids moving money without fresh approval and restricts the agent from asking for sensitive account secrets.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 22, 2026, 10:10 PM