The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill connects to https://api.clinpgx.org/v1/ to retrieve clinical data regarding gene-drug interactions and pharmacogenomic guidelines. This is the primary intended function of the skill.
[DATA_EXFILTRATION]: No sensitive local data or credentials are accessed or transmitted. The network operations are limited to pharmacogenomic queries and retrieving JSON responses from the database API.
[COMMAND_EXECUTION]: The provided Python script scripts/query_clinpgx.py uses the requests library for HTTP GET calls and does not utilize any dangerous shell commands, subprocesses, or arbitrary execution functions.
[PROMPT_INJECTION]: The skill instructions contain no patterns designed to override agent safety guidelines, bypass constraints, or extract system prompts.
[CREDENTIALS_UNSAFE]: The API does not require authentication for basic access as documented, and no hardcoded API keys or secrets were found in the provided files.
[SAFE]: The skill follows best practices for API interaction, including rate limiting (2 requests per second) and exponential backoff for error handling. A caching mechanism is provided to reduce network load, which writes data to local JSON files but does not represent a significant security risk in this context.

clinpgx-database