The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill integrates with external data sources which could potentially contain adversarial instructions designed to influence the agent's behavior.
Ingestion points: The dev server in src/dev/server.ts proxies requests to an external API (OPENEVIDENCE_BASE_URL) and the mock server in src/dev/mocks.ts provides clinical decision support data (answers and citations).
Boundary markers: No specific delimiters or instructions are provided to the agent to treat the API or mock data as untrusted content.
Capability inventory: The skill is granted Read, Write, and Edit tool permissions. The server scripts utilize the express framework and axios for networking.
Sanitization: There is no evidence of sanitization, validation, or escaping of the clinical data retrieved from the API or mock routes before it is presented to the agent context.
[EXTERNAL_DOWNLOADS]: The skill configuration guides the user to install several external dependencies.
Fetches express, axios, dotenv, tsx, typescript, @types/node, vitest, supertest, and @types/express from the official npm registry.
[COMMAND_EXECUTION]: The environment setup and testing workflow involve executing shell commands.
Includes standard commands for file management (cp .env.example .env) and package management (npm install).

openevidence-local-dev-loop