Brain Researcher

Brain Researcher

AI-assisted research infrastructure for neuroimaging

Turn neuroimaging questions into evidence-linked plans, runnable workflows, and reviewable scientific claims.

datasets
1,600+
tool specs
2,000+
MCP tools
87
workflows
60+

Tested across Claude, Codex, Gemini, GLM, DeepSeek, Kimi, and Qwen.

1

Install Claude Code

npm install -g @anthropic-ai/claude-code

npm install -g @anthropic-ai/claude-code
Get Claude Code
2

Get your token

Sign in to mint your personal MCP token (a one-time secret).

Log inCreate account
3

Configure MCP

Set the token in your shell first. Then keep the MCP config as a separate file.

Shell profile only

Add this to ~/.bashrc, ~/.zshrc, or ~/.profile. Start Claude Code from an environment that inherits BR_MCP_TOKEN. Do not paste these export lines into .mcp.json.

# Brain Researcher MCP (add to ~/.bashrc, ~/.zshrc, or ~/.profile)
export BR_MCP_TOKEN="brk_<kid>.<secret>"
export BR_MCP_AUTH_HEADER="Bearer ${BR_MCP_TOKEN}"
.mcp.json only

Paste only this config into .mcp.json. It reads BR_MCP_TOKEN from your client environment.

{
  "mcpServers": {
    "brain-researcher": {
      "type": "http",
      "url": "https://brain-researcher.com/mcp",
      "headers": {
        "Authorization": "Bearer ${BR_MCP_TOKEN}",
        "Accept": "application/json, text/event-stream"
      }
    }
  }
}
Try it
Claude alone vs. with Brain Researcher

Now, just ask.

Same question, same agent. Brain Researcher adds dataset discovery and an inspectable recipe for local execution.

your coding agent
›

Claudeno tools
Claude + Brain ResearcherMCP
Streaming demo…
Set up MCP
Paper & code
Open & reproducible

Read the paper, run the code

Paper

Bringing analytic rigor to agentic AI for science: The Brain Researcher platform for neuroimaging data analysis

arXiv:2608.19902 · cs.AI

AI agents can execute scientific analyses, but an analytic output becomes a defensible claim only after alternatives are weighed and the claim is limited to what the evidence supports. Agents may reproduce failures including selective analysis, premature declarations of success and optimization of imperfect criteria. We present Brain Researcher, an agentic research harness operating in a neuroimaging researcher's computational environment under rules for admissible analyses, required checks and claim scope. In benchmarks, Brain Researcher increased first-choice tool-selection accuracy across seven models by 70.2 percentage points (23.3% without it versus 93.6% with it) and verifiable grounding from 4.6% to 22.0%. In collaborator-led and self-evolving studies, multiverse analyses exposed analytic-choice sensitivity, and scientific review classified claims as accepted, qualified, revised, blocked, rejected or deferred. By linking decisions to evidence and provenance, Brain Researcher embeds methodological judgment within the workflow, not after it.

Read on arXiv
Code

Brain Researcher

Open-source release

The full platform: Next.js web UI, FastAPI orchestrator + agent, 87 public MCP tools, 2,000+ registry tool specs, and the Neo4j knowledge graph. We're open-sourcing the shipped code and supported workflow paths so you can inspect and adapt them in your own environment.

PythonTypeScriptNext.jsFastAPINeo4jMCP
On GitHub
Demos
See it in action

Explore worked demos

Curated use cases with real evidence bundles, reports, and handoff context. Open one to walk the full research episode.

Browse all demos
Back to top