The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes echo to verify that required environment variables such as SCALEKIT_CLIENT_ID and SCALEKIT_CLIENT_SECRET are present.
[COMMAND_EXECUTION]: Uses curl to perform a test authentication request against the Scalekit OAuth token endpoint to confirm credential validity.
[COMMAND_EXECUTION]: Utilizes ripgrep (rg) to search the project codebase for patterns associated with hardcoded secrets (e.g., skc_), which is an intended security audit feature to prevent accidental credential leakage.
[DATA_EXFILTRATION]: Transmits client credentials to a remote endpoint defined by $SCALEKIT_ENVIRONMENT_URL. As the skill identifies this as the vendor's official production environment (*.scalekit.com), this constitutes standard functionality for the product's OAuth verification process.
[PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection because it ingests and processes raw data from the local file system during its secret-scanning phase.
Ingestion points: Output from the ripgrep command scanning the project repository (SKILL.md).
Boundary markers: None; the agent is not instructed to disregard instructions found within scanned project files.
Capability inventory: File system access, shell command execution (curl, echo, rg), and interaction with the Scalekit SDK tools.
Sanitization: None; the agent processes the raw text returned from the file system search.

production-readiness-agentkit