Skip to main content
Glama
bharatvansh

gemini-deep-research-mcp

by bharatvansh
OverviewSchemaSourceRelated ServersDiscussionsAdmin
Python
Remote

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault
GEMINI_API_KEYYesYour Gemini API key from https://aistudio.google.com/apikey
GEMINI_DEEP_RESEARCH_AGENTNoModel to usedeep-research-pro-preview-12-2025

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

CapabilityDetails
tools
{
  "listChanged": false
}
prompts
{
  "listChanged": false
}
resources
{
  "subscribe": false,
  "listChanged": false
}
experimental
{}

Tools

Functions exposed to the LLM to take actions

NameDescription
start_deep_researchA

Initiates a deep, multi-step web research job in the background using Google's Deep Research Agent. Returns a job_id, which you can use to check the status of completion using check_deep_research(job_id=...).

Parameters:

  • prompt: The comprehensive research question or topic to investigate (required)

Returns:

  • job_id: Unique tracking ID for the research job

  • status: Initial job state (typically 'in_progress')

check_deep_researchA

Checks the status of a Deep Research job using its job_id and returns the complete report once finished.

Parameters:

  • job_id: The tracking ID returned by start_deep_research (required)

  • include_citations: Whether to include source URLs in the report (default: true)

Returns:

  • job_id: The tracking ID of the research job

  • status: Current job state ('in_progress', 'completed', 'failed', or 'cancelled')

  • report_text: The synthesized research report when completed

  • uptime: Elapsed time while the job is in progress, when available

  • error: Failure details, when the API provides them

Prompts

Interactive templates invoked by user choice

NameDescription

No prompts

Resources

Contextual data attached and managed by the client

NameDescription

No resources

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

A4.1/5.0

Scored across 2 tools

Disambiguation5/5

The two tools have completely distinct responsibilities: one initiates a research job accordion to a prompt, and the other polls for status and retrieves the report. There is no overlap in functionality or ambiguity about which tool to use in any given situation.

Naming Consistency5/5

Both tools follow a consistent verb_noun pattern: 'start_deep_research' and 'check_deep_research'. The action (start/check) clearly precedes the domain (deep_research), making the naming predictable and intuitive.

Tool Count3/5

With only two tools, the surface is minimal but matches the narrow scope of managing a deep research job. However, per the calibration this falls in the borderline category (1-2 tools). It is slightly thin, though not unreasonable for such a specific workflow.

Completeness4/5

The lifecycle is essentially complete: start a job and retrieve the final report. The absence of a cancel or list tools is a minor gap, but the core workflow (initiate and poll) is fully covered from the agent's perspective.

Maintenance

ActivityMaintained
ResponsivenessNo issues

Resources