The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides instructions for deploying applications using a local Node.js script located at .codex/skills/taruvi-frontend-worker-deploy/scripts/deploy-frontend-worker.mjs. This script is part of the vendor's deployment pipeline and is used to automate the build and upload process to the Taruvi cloud platform.\n- [CREDENTIALS_UNSAFE]: Authentication for the Taruvi platform is managed through the TARUVI_API_KEY environment variable. The skill correctly advises storing these credentials in project-level .env files and includes precautions against logging the API key in system logs or output.\n- [EXTERNAL_DOWNLOADS]: Deployment operations involve communication with api.taruvi.cloud. This is the official API endpoint for the vendor's platform and is required for the skill to perform its intended tasks.\n- [PROMPT_INJECTION]: The skill is designed to analyze existing project code to provide contextual recommendations. This introduces a surface for indirect prompt injection if a project contains malicious instructions intended to mislead the AI, though this is a standard risk for development tools and is mitigated by the vendor-focused nature of the skill.\n
Ingestion points: Reads project files (e.g., .env, existing source code) to determine implementation requirements.\n
Boundary markers: None explicitly implemented for delimiting code provided by the user or project files.\n
Capability inventory: Executes local Node.js scripts for deployment and generates source code for the user.\n
Sanitization: Relies on general safety guardrails; no specific sanitization logic is described for the analyzed project data.

taruvi-app-developer