The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the installation of the notebooklm-mcp-cli package from PyPI. This is a third-party dependency that is not associated with a known trusted organization.
[COMMAND_EXECUTION]: The instructions involve running shell commands to install software, log into services, and locate executable paths (pip install, nlm login, which). It also requires modifying the opencode.json configuration file in the user's home directory (~/.config/opencode/opencode.json) to enable the MCP server functionality.
[PROMPT_INJECTION]: By integrating with Google NotebookLM, the agent processes external data which may contain untrusted content. This creates a surface for indirect prompt injection if the notebooks contain instructions intended to hijack the agent's logic.
Ingestion points: Notebook content retrieved from Google NotebookLM via the notebooklm MCP tool defined in SKILL.md.
Boundary markers: No boundary markers or specific "ignore" instructions are provided to the agent when processing retrieved notebook data.
Capability inventory: The agent is granted the ability to list, read, and write notebooks using the nlm command-line tool.
Sanitization: There is no evidence of content sanitization or validation before the agent processes the retrieved data.

opencode-notebooklm