The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill implements patterns for ingesting and processing external data from API responses and HTTP requests, creating a surface for indirect prompt injection.
Ingestion points: Data is ingested via the Axios client in apps/web-app/src/utils/apiClient.js and through request body/query parameters in the Express controllers, such as services/user-service/src/controllers/user.controller.js.
Boundary markers: The provided code patterns do not include explicit delimiters or specific instructions for the agent to isolate untrusted content from the rest of the prompt context.
Capability inventory: The skill patterns include capabilities for database persistence using Mongoose (user.save() in user.controller.js), network operations via Axios (apiClient.js), and the use of the Bash and Task tools as defined in the skill's allowed tools.
Sanitization: The patterns implement input validation using express-validator (e.g., in services/user-service/src/routes/users.js) to normalize and validate email addresses and string lengths before data is processed by the controllers.

mern-patterns