The Agent Skills Directory

[SAFE]: The skill provides legitimate implementation patterns for API rate-limiting and does not contain any malicious instructions or hidden behaviors.
[SAFE]: The code snippets correctly utilize environment variables (process.env.NAVAN_ACCESS_TOKEN) for authentication, avoiding hardcoded secrets and following standard security practices.
[COMMAND_EXECUTION]: Includes a simple curl command in the examples section intended for manual testing of API response headers. This command is transparent and performs no unauthorized actions.
[INDIRECT_PROMPT_INJECTION]: The skill implements logic to process HTTP response headers (X-RateLimit-*, Retry-After) from the Navan API. This data ingestion is handled safely through integer parsing and is restricted to controlling request timing and logging, posing no risk of instruction override or capability abuse.

navan-rate-limits