The Agent Skills Directory

[COMMAND_EXECUTION]: The skill guides the agent to use the nlm command-line tool for managing NotebookLM resources. This includes operations for creating notebooks, adding sources, and generating AI-driven content like podcasts and reports.
[EXTERNAL_DOWNLOADS]: The documentation references the notebooklm-mcp-cli package for installation. This is the primary tool for the skill and is authored by the vendor.
[PROMPT_INJECTION]: No malicious injection patterns or safety bypass attempts were found. The instructions reinforce agent safety by requiring explicit user confirmation before any delete or sync commands.
[DATA_EXFILTRATION]: Authentication is handled via standard browser-based cookie extraction and local profile management. No evidence of sensitive data being sent to non-whitelisted or unauthorized domains was observed.

nlm-skill