The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill provides templates for processing external data such as headers and request payloads.
Ingestion points: The code snippets process req.headers['idempotency-key'] and payload variables.
Boundary markers: Not applicable as the skill provides application-level code templates rather than agent-facing instructions.
Capability inventory: The templates perform standard database operations (INSERT, UPDATE, DELETE) and cache operations (get, setEx).
Sanitization: The code demonstrates security best practices by using parameterized SQL queries (e.g., $1, $2) and SHA-256 hashing to verify request integrity.
[COMMAND_EXECUTION]: The provided code involves standard backend application logic for interacting with databases and caching layers. No risky subprocess calls, shell command execution, or unsafe system operations were detected.
[DATA_EXPOSURE]: The skill demonstrates how to store and retrieve API responses for idempotent processing. It includes best practices for data lifecycle management, such as implementing TTL (Time-To-Live) and periodic cleanup jobs to prevent storage bloat.

idempotency-handling