The Agent Skills Directory

[DATA_EXFILTRATION] (SAFE): The environment-config.md file contains examples of database connection strings (postgresql+asyncpg://user:password@host:5432/dbname) and API keys, but these are explicitly labeled as placeholders or development-only values (e.g., dev-secret-key-not-for-production). The skill correctly instructs users to store real secrets in AWS Secrets Manager or GitHub Secrets.
[COMMAND_EXECUTION] (SAFE): The migration-dry-run.sh script executes shell commands to perform PostgreSQL operations and Alembic migrations. These operations are performed on a temporary clone of the database (_migration_test) created by the script, which is a standard and safe practice for validating migrations before applying them to production.
[EXTERNAL_DOWNLOADS] (SAFE): The health-check.py and smoke-test.sh scripts make network requests using standard libraries (urllib.request) and system utilities (curl). These requests are restricted to the base URL provided by the user and are used exclusively for monitoring service availability and status codes.
[CREDENTIALS_UNSAFE] (SAFE): No hardcoded production credentials were found. All sensitive configuration is handled via environment variables, and the documentation includes clear warnings against committing .env files to version control.
[PROMPT_INJECTION] (SAFE): No instructions attempting to bypass safety filters or override agent behavior were detected in the documentation or script comments.

deployment-pipeline