Schema | Gemini Notebook by Chryzalide (agy-gemini-notebook-plugin)

Gemini Notebook by Chryzalide (agy-gemini-notebook-plugin)

Overview Schema Source Related Servers Discussions Admin

Server Configuration

Describes the environment variables required to run the server.

Name	Required	Description	Default
`GCP_LOCATION`	No	Google Cloud location for Engine 1. Supported values include global, us, or eu.	global
`AGY_WORKSPACE_DIR`	No	Active workspace directory for the MCP server, typically set to the IDE workspace folder.
`GCP_PROJECT_NUMBER`	No	Google Cloud project number for Engine 1 official NotebookLM Enterprise API. If not set, the server falls back to Engine 2 RPC bridge.
`GCP_ENDPOINT_LOCATION`	No	Google Cloud endpoint location for Engine 1. Supported values include global, us, or eu.	global

Instructions

Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.

This server publishes no instructions, or was last inspected before Glama recorded them.

Capabilities

Features and capabilities supported by this server

Protocol revision2025-11-25

Capability	Details
`tools`	{ "listChanged": false }
`prompts`	{ "listChanged": false }
`resources`	{ "subscribe": false, "listChanged": false }

Tools

Functions exposed to the LLM to take actions

Name	Description
notebook_statusB	Check current workspace Gemini Notebook binding and authentication status for both engines.
notebook_loginA	Connect your Gemini Notebook session without Playwright sign-in automation on `accounts.google.com`. - Default (`browser_cookies='auto'` or `'chrome'`, `'brave'`, `'firefox'`, `'edge'`, `'chromium'`): Imports your already-signed-in `notebooklm.google.com` session directly from your installed browser in ~1 second (Engine 2 — unofficial community RPC bridge). - `gcloud=True`: Runs official `gcloud auth login` OAuth 2.0 flow for Google Cloud Discovery Engine `v1alpha` (Engine 1 — 100% official Google Cloud API). - `open_system_browser=True`: Opens `https://notebooklm.google.com/` in your genuine OS default browser so you can verify you are signed in before connecting.
notebook_attachC	Attach a Gemini Notebook to the active workspace (`.agents/notebook.json`). Supports optional `alias` so multiple notebooks can be registered and invoked by short name (e.g. `alias="architecture"` or `alias="research"`).
notebook_invokeB	Invoke a Gemini Notebook by ID, workspace alias, or title to ground the current task. - If `question` is provided: executes a grounded query against the target notebook and returns the synthesized answer, inline citations, and follow-up suggestions. - If `question` is omitted: fetches the notebook's metadata, AI summary, suggested topics, and source list so the agent can immediately inspect and use the notebook context.
notebook_listC	List Gemini Notebooks via official GCP v1alpha REST API or RPC bridge, with optional title filter.
notebook_getC	Get metadata, AI summary, and list of sources for a Gemini Notebook (Dual-Engine).
notebook_createC	Create a new Gemini Notebook via GCP v1alpha REST API or RPC bridge and optionally attach it.
notebook_deleteC	Delete one or more Gemini Notebooks via GCP v1alpha `batchDelete` or RPC bridge.
notebook_renameC	Rename a Gemini Notebook via the RPC bridge and update `.agents/notebook.json` if attached.
notebook_shareC	Share a Gemini Notebook with an email (GCP Enterprise IAM) or toggle public link sharing (RPC).
notebook_sync_filesC	Upload local workspace files (.md, .pdf, .txt, .docx, .pptx, .xlsx, audio, images) to a Gemini Notebook.
notebook_add_sourceA	Add a text, web URL, YouTube video, or Google Drive document to a Gemini Notebook (Dual-Engine). Args: source_type: One of 'text', 'web', 'youtube', or 'gdrive'. value: Raw text content, URL, YouTube URL, or Google Drive documentId. title: Display name for the source. mime_type: Used only when `source_type == 'gdrive'`. notebook_id: Target notebook ID or workspace alias (defaults to workspace-attached notebook). wait: When using RPC bridge, whether to wait for source indexing to complete. engine: 'auto', 'gcp_rest', or 'rpc_bridge'. workspace_dir: Optional workspace directory override.
notebook_delete_sourcesC	Delete one or more sources from a Gemini Notebook (Dual-Engine).
notebook_search_sourcesC	Search indexed passages across a Gemini Notebook's sources and return ranked chunks.
notebook_get_source_textC	Fetch the full indexed text (and optional AI summary/keywords guide) of a specific source.
notebook_queryB	Ask a grounded question against a Gemini Notebook and receive answer + inline citations. Automatically filters to `active_source_ids` in `.agents/notebook.json` if `source_ids` is not explicitly passed. Supports multi-turn follow-ups via `conversation_id` and persisting the cited answer as a notebook note via `save_as_note=True`.
notebook_researchA	Launch a Fast Research, Deep Research, or Source Discovery session on a Gemini Notebook. Allows the agent harness to pilot NotebookLM as an autonomous research engine: - Set `create_notebook_title` to spin up a brand-new notebook before launching research. - Set `mode='fast'` for rapid web/Drive search (~10-30s). - Set `mode='deep'` for exhaustive multi-step Deep Research with a synthesized Markdown report. - Set `mode='discover'` (or `'curious'`) for instant synchronous web source discovery. - Set `source='web'` or `source='drive'`. - When `wait=True` and `import_sources=True`, discovered sources are automatically imported into the notebook so you can immediately query or generate artifacts from them.
notebook_research_statusC	Check status, wait for completion, import sources from, or cancel a Fast/Deep Research run.
notebook_create_audioC	Generate an Audio Overview podcast on the notebook via GCP v1alpha REST or RPC bridge.
notebook_generate_artifactC	Generate a NotebookLM Studio Artifact (`report`, `study_guide`, `briefing_doc`, `blog_post`, `mind_map`, `quiz`, `flashcards`, `slide_deck`, `data_table`, `infographic`, `video`, `audio`).
notebook_notesC	Manage persistent notes inside a Gemini Notebook (`action`: 'list', 'get', 'create', 'update', 'delete').

Prompts

Interactive templates invoked by user choice

Name	Description
`invoke_notebook`	Invoke a Gemini Notebook by ID, alias, or title to ground the current coding or research task.
`deep_research_notebook`	Create or select a Gemini Notebook, run Deep Research on a topic, import cited sources, and synthesize findings.

Resources

Contextual data attached and managed by the client

Name	Description
`notebook_workspace_status`	Active workspace Gemini Notebook binding and Dual-Engine authentication status.
`active_workspace_notebook`	Metadata, AI summary, and source list for the workspace's active Gemini Notebook.

Deploy Server

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

TDQS

C2.9/5.0

Scored 2026-10-25 13:50 across 21 tools

Disambiguation3/5

Several tools overlap in practice: `notebook_get` and `notebook_invoke` without a question both return metadata/summary, and `notebook_query` versus `notebook_invoke` with a question are functionally near-identical grounded Q&A paths. `notebook_add_source` and `notebook_sync_files` also both ingest content, and `notebook_notes` bundles five actions into one tool. Descriptions do disambiguate somewhat, but boundary choice requires careful reading.

Naming Consistency4/5

All 21 tools use a consistent snake_case `notebook_*` prefix, which makes the family instantly recognizable. Minor deviation: some names are verb-first (create, delete, rename, list, query) while others are noun/state-like (`notebook_status`, `notebook_notes`, `notebook_research_status`), but the pattern remains readable and predictable.

Tool Count3/5

21 tools is on the heavy side for a single-product integration, and the dual-engine (GCP REST vs RPC bridge) design is baked into tool parameters rather than split into separate surfaces. Most tools map to a real operation, but the set sits at the upper edge of what an agent can hold comfortably in one context.

Completeness4/5

Coverage is broad: notebook lifecycle, sources, notes, grounded search/query, research sessions, sharing, file sync, and artifact generation. Gaps exist around updating notebook metadata (only `notebook_rename`), editing/replacing sources, and retrieving or listing generated artifacts after creation. These are workarounds-able but notable.

Maintenance

ActivityMaintained

ResponsivenessNo issues