The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides instructions and configurations to install the official Google Generative AI SDK (@google/genai) and TypeScript from the public npm registry.
[DATA_EXFILTRATION]: Communication is restricted to official Google API endpoints (generativelanguage.googleapis.com) for the purpose of generating text embeddings and LLM responses as described in the skill's primary functionality.
[INDIRECT_PROMPT_INJECTION]: As a guide for Retrieval-Augmented Generation (RAG), the skill naturally processes external document content and user queries. It includes documentation on best practices for handling this data surface, such as boundary markers and sanitization, to mitigate risks inherent to processing untrusted input in LLM applications.
[COMMAND_EXECUTION]: The skill includes a standard maintenance script (scripts/check-versions.sh) that uses npm view to verify package versions. This is a common developer utility and does not perform any unauthorized or dangerous system operations.
[CREDENTIALS_UNSAFE]: The skill correctly instructs users to manage API keys via environment variables or secrets (e.g., GEMINI_API_KEY) rather than hardcoding credentials into source code.

google-gemini-embeddings