The Agent Skills Directory

[SAFE]: The skill provides documentation for the official Runpod Python SDK. No malicious patterns were detected.
[INDIRECT_PROMPT_INJECTION]: The skill defines a toolchain for processing and executing user-authored Python code locally and remotely. Ingestion points include project source files and configuration. The capability inventory includes local execution, remote deployment, and file creation. Boundary markers are defined by the project directory structure. Sanitization is recommended through the use of .runpodignore to exclude sensitive files.
[CREDENTIALS_UNSAFE]: The documentation provides instructions on managing authentication via environment variables and configuration files, and warns that API keys are injected into worker environments by default.

runpod-flash