Enables an AI client to perform autonomous research on complex problems by verifying claims, identifying assumptions, and generating evidence-backed answers.
Also known as thinkerwave-ai/thinkerwave, thinkerwave-mcp
- Stars
- 0
- Tools
- —
- Downloads, 30 days
- 348
- Latest version
- 0.1.2
- Last commit
- Oct 20, 2026
- License
- Apache-2.0
What is ThinkerWave MCP?
ThinkerWave enables an AI client to perform autonomous research on complex problems by verifying claims, identifying assumptions, and generating evidence-backed answers.
Summary generated from the project's README
About ThinkerWave MCP server
ThinkerWave provides an autonomous research agent designed for complex problems that lack a single correct answer. It is built for tasks like market entry analysis, due diligence, credit risk stress tests, and drug program selection. Instead of providing a simple response, the system researches, calculates data, and iterates through multiple rounds to refine its findings. Users receive a final answer accompanied by a confidence score and specific evidence categorized as verified, an assumption, or a guess.
Developers can integrate this capability into their workflows using the Python SDK, the tw command line interface, or as an MCP server for clients like Claude Code and Cursor. While the engine runs in the ThinkerWave cloud, you connect to it using a THINKERWAVE_API_KEY. The server supports running a specific number of research rounds to ensure quality. Free runs are available for every account, subject to defined usage limits.
What you can do with it
- 1Perform complex market researchHand a difficult business question to your coding assistant to receive an evidence-backed answer.
- 2Validate business assumptionsProvide known facts to the agent so it can research and verify claims through multiple iterations.
Written from the project's README. Every use case is checked against a quote from it.
How to install ThinkerWave MCP
Formats verified 10 Nov 2026Add to CursorAdd to GooseInstall ThinkerWave MCP in Claude Code
$ claude mcp add --env 'THINKERWAVE_API_KEY=<YOUR_THINKERWAVE_API_KEY>' thinkerwave -- uvx thinkerwave-mcpInstall ThinkerWave MCP in Cursor
~/.cursor/mcp.json
{
"mcpServers": {
"thinkerwave": {
"command": "uvx",
"args": [
"thinkerwave-mcp"
],
"env": {
"THINKERWAVE_API_KEY": "<YOUR_THINKERWAVE_API_KEY>"
}
}
}
}Install ThinkerWave MCP in VS Code
.vscode/mcp.json
{
"servers": {
"thinkerwave": {
"type": "stdio",
"command": "uvx",
"args": [
"thinkerwave-mcp"
],
"env": {
"THINKERWAVE_API_KEY": "<YOUR_THINKERWAVE_API_KEY>"
}
}
}
}Install ThinkerWave MCP in VS Code CLI
$ code --add-mcp '{"name":"thinkerwave","type":"stdio","command":"uvx","args":["thinkerwave-mcp"],"env":{"THINKERWAVE_API_KEY":"<YOUR_THINKERWAVE_API_KEY>"}}'Install ThinkerWave MCP in Claude Desktop
claude_desktop_config.json
{
"mcpServers": {
"thinkerwave": {
"command": "uvx",
"args": [
"thinkerwave-mcp"
],
"env": {
"THINKERWAVE_API_KEY": "<YOUR_THINKERWAVE_API_KEY>"
}
}
}
}Install ThinkerWave MCP in Codex
$ codex mcp add thinkerwave --env 'THINKERWAVE_API_KEY=<YOUR_THINKERWAVE_API_KEY>' -- uvx thinkerwave-mcpInstall ThinkerWave MCP in Codex config.toml
~/.codex/config.toml
[mcp_servers.thinkerwave] command = "uvx" args = ["thinkerwave-mcp"] [mcp_servers.thinkerwave.env] THINKERWAVE_API_KEY = "<YOUR_THINKERWAVE_API_KEY>"
Install ThinkerWave MCP in Gemini CLI
$ gemini mcp add -s user -e 'THINKERWAVE_API_KEY=<YOUR_THINKERWAVE_API_KEY>' thinkerwave uvx -- thinkerwave-mcpInstall ThinkerWave MCP in Gemini settings.json
~/.gemini/settings.json
{
"mcpServers": {
"thinkerwave": {
"command": "uvx",
"args": [
"thinkerwave-mcp"
],
"env": {
"THINKERWAVE_API_KEY": "<YOUR_THINKERWAVE_API_KEY>"
}
}
}
}Configuration
THINKERWAVE_API_KEYRequiredSecretYour ThinkerWave API key (thinkerwave.ai/app → Settings → API keys)
Automated checks
Ran Nov 14, 2026Full report →Passed our automated checks. That is not a security audit: read the code and give the server the narrowest access it needs. How checks work
- Source repository linked and reachablethinkerwave-ai/thinkerwave is public and reachablegithub.com/thinkerwave-ai/thinkerwave ↗Pass
- Package names this repository as its sourcethinkerwave-mcp names thinkerwave-ai/thinkerwave as sourcepypi.org/project/thinkerwave-mcp/ ↗Pass
- Build provenance (Sigstore attestation)Sigstore attestation for thinkerwave-mcppypi.org/project/thinkerwave-mcp/ ↗Pass
- Open source license (OSI)Apache-2.0spdx.org/licenses/Apache-2.0.html ↗Pass
- Commit or release in the last yearLast commit or release 2026-10-20github.com/thinkerwave-ai/thinkerwave ↗Pass
- OpenSSF ScorecardNot checked
- Package name not close to a popular packageNo popular package one edit away from thinkerwave-mcpPass
- No critical advisory in the package or its dependencies33 resolved dependencies checked; no critical advisoryosv.dev ↗Pass
- No hidden instructions to the model0 tool descriptions and the README read; no override phrases, credential paths or hidden charactersPass
- Destructive tools are labelledNot checked
- Remote endpoint answers the MCP handshakeNot checked
README
View on GitHubThinkerWave
An autonomous research agent for problems without a single right answer — credit risk stress tests, pricing and churn, market entry, due diligence, drug program selection.
You give it the problem and the facts you're sure of. It researches, runs the numbers, checks its own work and tries again, round after round, keeping the best answer. You get the answer, how confident it is, and the evidence behind each claim — verified, assumption or guess — plus any charts and tables it produced.

pip install thinkerwave
export THINKERWAVE_API_KEY=tw_live_… # create one at thinkerwave.ai/app → Settings → API keys
tw run "Should we raise prices 15%, 8%, or hold? 1,200 customers at \$85/month, 3.1% monthly churn." \
--given "last 8% increase spiked churn to 4.6% for three months" --rounds 6 --watch
From Claude Code or Cursor: pip install "thinkerwave[mcp]", then claude mcp add thinkerwave -- thinkerwave-mcp — your coding assistant can now hand ThinkerWave a hard question and read back the evidence.
From Python: see examples/.
What's in this repo: the Python SDK, the tw command line, an MCP server, the API reference and the answer/evidence format. ThinkerWave itself runs in ThinkerWave's cloud; this repository is the way in, not the engine.
Free runs are included with every account. Limits: docs/limits.md. Found a problem it got wrong? Open an issue with the run id.
ThinkerWave MCP questions
How do I add ThinkerWave MCP to Claude Code?
Run claude mcp add --env 'THINKERWAVE_API_KEY=<YOUR_THINKERWAVE_API_KEY>' thinkerwave -- uvx thinkerwave-mcp in your terminal. Add --scope user to make it available in every project.
How do I add ThinkerWave MCP to Cursor?
Open ~/.cursor/mcp.json and add the server under mcpServers, using the Cursor config shown in the install section above.
How does ThinkerWave label its findings?
The agent labels every claim as either verified, an assumption, or a guess to provide transparency in its reasoning.
Where does the ThinkerWave engine actually run?
The ThinkerWave engine runs in the cloud; this repository provides the interface and SDK to access it.
Can I use ThinkerWave with Claude Code or Cursor?
Yes, you can install the MCP version and add it to your coding assistant to hand it hard questions.