The Agent Skills Directory

[COMMAND_EXECUTION]: The skill facilitates the execution of local binaries including the nlm CLI and notebooklm-mcp server to interact with Google NotebookLM services. It also utilizes shell scripts for environment verification and configuration.
[DYNAMIC_EXECUTION]: The script scripts/setup_notebooklm_mcp.sh executes an embedded Python script via stdin to perform structured JSON patching of agent configuration files. This method is used to ensure cross-platform compatibility without requiring external JSON manipulation tools.
[PERSISTENCE]: The skill provides automated configuration utilities that modify persistent agent configuration files (such as ~/.claude.json, ~/.cursor/mcp.json, and other platform-specific paths) to ensure the NotebookLM MCP server is registered and available across sessions. This is standard behavior for MCP-native skills.
[INDIRECT_PROMPT_INJECTION]: The skill provides an interface to ingest and process external data from Google NotebookLM, which serves as a potential vector for indirect prompt injection.
Ingestion points: Data is retrieved from NotebookLM notebooks, source documents, and chat history via the tool's CLI and MCP interfaces as described in SKILL.md and references/auth-guide.md.
Boundary markers: The instructions do not define specific delimiters or 'ignore' instructions for the data ingested from NotebookLM.
Capability inventory: The skill possesses the capability to execute shell commands (nlm), write to local configuration files (scripts/setup_notebooklm_mcp.sh), and perform network operations via the nlm binary.
Sanitization: No explicit sanitization or filtering of the content retrieved from Google services is implemented before it is presented to the agent context.

notebooklm-mcp