The Agent Skills Directory

[PROMPT_INJECTION]: The skill establishes an indirect prompt injection surface as it is designed to fetch and process data from external, third-party APIs hosted on a marketplace. \n
Ingestion points: Response data from arbitrary APIs called via the Nightmarket proxy (e.g., https://nightmarket.ai/api/x402/<endpoint_id>). \n
Boundary markers: The instructions do not define delimiters or negative constraints to prevent the agent from following instructions embedded in third-party API responses. \n
Capability inventory: The agent has the capability to make subsequent HTTP requests and authorize payments via CrowPay based on data received. \n
Sanitization: No sanitization or validation logic is specified for the external content processed by the agent. \n- [COMMAND_EXECUTION]: The skill instructions and reference documents recommend using shell commands for setup and execution. \n
Evidence: Commands such as npx skills add elizaos/eliza --skill nightmarket and npx skills add elizaos/eliza --skill crow are used for installation. \n- [EXTERNAL_DOWNLOADS]: The skill downloads external packages and communicates with external API services. \n
Evidence: The MCP configuration in references/mcp.md executes npx -y nightmarket-mcp, which downloads and runs the Nightmarket MCP package from the npm registry. The skill also makes network requests to nightmarket.ai and api.crowpay.ai for its core functionality.

nightmarket