Skip to main content
Glama
yashavsarmal30

Multi-Agent Deep Researcher MCP

by yashavsarmal30
OverviewSchemaSourceRelated ServersDiscussionsAdmin
JavaScript
Hybrid

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault
PORTNoWeb API & UI port.5000
LLM_MODELNoCustom model name (e.g., gpt-4o-mini, llama-3.3-70b-versatile, deepseek-r1:7b).
GROQ_API_KEYNoGroq Cloud API key. Optional.
LLM_PROVIDERNoPreferred LLM provider. Options: openai, groq, anthropic, gemini, deepseek, ollama.
GEMINI_API_KEYNoGoogle AI Studio Gemini API key. Optional.
LINKUP_API_KEYNoLinkUp Search API key. Optional; falls back to DuckDuckGo.
OPENAI_API_KEYNoOpenAI API key. Optional.
OLLAMA_BASE_URLNoOllama local API base URL.http://localhost:11434
DEEPSEEK_API_KEYNoDeepSeek Platform API key. Optional.
ANTHROPIC_API_KEYNoAnthropic Claude API key. Optional.

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
deep_researchA

Run an autonomous multi-agent deep research investigation on a topic.

Coordinates a Web Researcher, Research Analyst, and Technical Writer
to gather live web data, synthesize insights, and produce a fully cited
Markdown report.

Args:
    query: The research topic, query, or question to investigate.
    depth: Thoroughness level - 'standard' for concise, 'deep' for exhaustive.
    search_engine: Search engine to use - 'auto' (preferred), 'linkup', or 'duckduckgo'.
    model: Optional model identifier (e.g., 'gpt-4o-mini', 'llama-3.3-70b-versatile').
    provider: Optional provider ('openai', 'groq', 'anthropic', 'gemini', 'ollama').

Returns:
    The complete publication-grade Markdown research report with source citations.
quick_searchA

Conduct a fast web search and return structured snippets with source URLs.

Args:
    query: The search term or question.
    max_results: Maximum number of links to return (default: 5, max: 15).
    search_engine: Search engine to use ('auto', 'linkup', or 'duckduckgo').

Returns:
    Formatted search results with titles, snippets, and URLs.
list_research_reportsA

List all previously generated research reports saved in the reports directory.

Returns:
    JSON list of available reports with filename, query, and creation date.
read_research_reportA

Retrieve and read a saved research report from disk.

Args:
    filename: The filename of the report (e.g., '20260306_quantum_computing.md').

Returns:
    The full text of the research report.

Prompts

Interactive templates invoked by user choice

NameDescription
deep_research_briefTemplate for initiating a structured deep research request.
competitive_analysisTemplate for a comparative research analysis between two entities or technologies.

Resources

Contextual data attached and managed by the client

NameDescription
research://status
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 4 tools

Disambiguation5/5

Each tool maps to a distinct task: generating a synthesized report, running a lightweight search, listing existing reports, and reading a specific report. Deep_research and quick_search are related but clearly separated by output depth and purpose.

Naming Consistency4/5

The names are clear and mostly follow a readable pattern, with list_research_reports and read_research_report using verb_noun construction. deep_research and quick_search break that pattern by leading with a modifier, but all names are concise snake_case and easy to predict.

Tool Count5/5

Four tools is a well-scoped size for a research server: one for investigation, one for quick lookup, and two for managing generated reports. No tool feels redundant or unnecessary.

Completeness4/5

The set covers the core research workflow and report retrieval end-to-end. A delete/remove report operation would make report lifecycle management more complete, but agents can still list, read, and generate reports without dead ends.

Maintenance

ActivityMaintained
ResponsivenessNo issues

Resources