brainstorm_orchestrateLABSCO SUMMARY
The value is in the disagreement, not the volume.
A single model asked for a design tends to produce one confident answer with no visible alternatives, and no pressure to justify scope. Running three roles against each other creates that pressure — the feasibility pass in particular exists to cut architectures that ignore a stated constraint. Two limits to know before adopting it: the LLM adapter is DeepSeek-only today, and the three roles are fixed, so custom perspectives are not yet available.
What it is
A multi-agent design engine exposed over MCP. Give it a task and it decomposes the requirement, fills in the parts you did not mention, has three role-based agents each produce a full proposal, makes them cross-review one another, forces a feasibility pass that cuts over-engineering, then scores everything on five dimensions and returns the top three.
What you get
brainstorm_orchestrateruns the whole pipeline from a task description;agentscontrols how many perspectives run, from a single fast pass up to the full three- Because a full run takes minutes, it is asynchronous:
brainstorm_statusreports progress by task ID andbrainstorm_resultreturns the finished output - Gap detection — the requirement list is audited before any design starts, and missing pieces such as security, deployment or migration are added
- A feasibility brake: an engineering role cuts proposals that exceed the stated constraints, and an over-built design is pushed down the ranking rather than quietly returned
- Scoring on correctness, completeness, feasibility, innovation and business alignment, with the reasoning attached to each score
- Output as a Markdown blueprint plus a self-contained HTML report;
collusion_blackboard_statusreports on shared-blackboard runs
Requirements
Python 3.10 or higher and a DeepSeek API key. The README's install line names collusion-mcp, which PyPI does not have; that is also the console script the project declares, at version 0.5.1. The key is read from DEEPSEEK_API_KEY, falling back to OPENAI_API_KEY then LLM_API_KEY, or it can go in a config.json. Clients launch the server module over stdio; an SSE mode is started with --sse --port 8020.
Setup effort
Build from source — git clone https://github.com/nideweilaiya/Collusion.git
USE CASES
01
Choosing an architecture with more than one option on the table
Get three independently written designs plus a ranking, rather than the first plan that sounded reasonable.
02
Catching what the requirement left out
Let the requirement audit add the security, deployment and migration work nobody mentioned in the brief.
03
Keeping a design inside its constraints
The feasibility pass removes the extra services a proposal added, and the over-built option ranks lower.
04
Checking on a long run without blocking
Poll by task ID while the agents work, then collect the report when it is finished.
4 tools
Listed in project docs. Your client may show a different list.
brainstorm_statusbrainstorm_resultcollusion_blackboard_statusBEFORE YOU INSTALL
What to know first
Tools you get
4 tools named in the project's documentation.
Requires
- Python 3.10+
- DEEPSEEK_API_KEY in your environment
INSTALLATION
Connect your client
You clone this one yourself.
Server setup
TERMINAL
git clone https://github.com/nideweilaiya/Collusion.git
cd Collusion
pip install -e .
export DEEPSEEK_API_KEY="<YOUR_DEEPSEEK_API_KEY>"MCP Clients
Claude Code
CLAUDE CODE · JSON
{
"mcpServers": {
"brainstorm": {
"command": "python",
"args": [
"<YOUR_PATH_TO_COLLUSION>/src/mcp_server.py",
"--stdio"
],
"env": {
"DEEPSEEK_API_KEY": "<YOUR_DEEPSEEK_API_KEY>"
}
}
}
}Trae
In the Collusion folder, start SSE mode:
TRAE · TERMINAL
python src/mcp_server.py --sse --port 8020Then add http://localhost:8020/sse to your MCP configuration.
