The Agent Skills Directory

[SAFE]: The skill provides documentation for a legitimate integration package within the LangChain ecosystem. No malicious patterns or behaviors were identified.
[EXTERNAL_DOWNLOADS]: The skill instructs the user to install the langchain-openrouter package from PyPI and references numerous other LangChain-related libraries. All packages mentioned are standard components or integrations of the well-known LangChain framework.
[CREDENTIALS_UNSAFE]: The documentation provides standard security practices for handling API keys, such as using the OPENROUTER_API_KEY environment variable. Code examples use clearly labeled placeholders like your_api_key or "..." for constructor parameters.
[INDIRECT_PROMPT_INJECTION]: The documented invoke and bind_tools interfaces allow for natural language processing and tool interaction. While this creates a capability surface for the agent, the skill does not contain any malicious prompts or vulnerabilities, and these features represent the intended functionality of the library.

langchain-openrouter