The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill requires a private key (INTEL_PRIVATE_KEY) for crypto wallet transactions on the Base network. It explicitly instructs users to provide this key in URLs (e.g., https://intel-mcp.asrai.me/mcp?key=0x<your_private_key>), environment variables, or JSON configuration files. Storing or transmitting raw private keys is a dangerous practice that can lead to total loss of funds if the key is intercepted or the tool is compromised.
[EXTERNAL_DOWNLOADS]: The installation and configuration instructions use npx -y intel-asrai-mcp, which fetches and executes a package from the npm registry at runtime.
[COMMAND_EXECUTION]: The skill configuration involves executing shell commands via npx to run the MCP server, which introduces a vector for executing code provided by the package author.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection.
Ingestion points: The intel_search tool retrieves content from external web sources, academic papers, and discussion forums.
Boundary markers: None identified in the provided instructions to separate untrusted web content from agent instructions.
Capability inventory: The skill uses the intel_search tool to synthesize answers from external data.
Sanitization: There is no evidence of sanitization or filtering of the retrieved web content before it is processed by the AI.

intel-asrai