Your AI Development Operating System

Drop into any project, run one command, and get scaffolding, comprehension, ideation, and quality gates — plus auto-learning that makes every session smarter than the last.

Get Started Free View on GitHub
$ pip install brainstormer
$ brainstormer init

  BrainStormer — The AI Development Operating System
  ==================================================

  Detecting project...          Next.js + TypeScript + Supabase
  Scaffolding project docs...   10 created, 0 existing
  Setting up comprehension...   CodeGlass ready
  Setting up ideation...        BrainStormer ready
  Setting up quality gates...   PALADIN ready
  Syncing to Obsidian...        Done

  Done! Project 'my-app' is ready.
  

For developers AND vibe coders

📌

Kernel

Scaffolds 10 project docs (CLAUDE.md, SPEC.md, ARCHITECTURE.md, etc.) with smart defaults based on your detected stack. 5-tier LLM routing included.

👁

Comprehension

Every code output gets a WHAT/HOW/WHERE/WHEN/WHY walkthrough. Auto-generates rules as patterns emerge. Your project gets smarter every session.

💡

Ideation

6-step idea multiplier: audience lock, core extraction, hook matching, calendar build, outline expansion, and skill deposit. One idea becomes dozens.

🛡

Quality

6-tier testing wall (Obvious to Human). Catches typos before you debug race conditions. Auto-bootstraps missing test infrastructure.

🧠
NEW

Self-Learning Rules

Auto-proposes rules from git diffs. Tracks confidence over time. Prunes stale rules. Your project's knowledge compounds with every commit.

🤖

374 Agents

Curated from 4 community sources. Smart surfacing shows relevant agents for your stack. Create custom agents and multi-agent pipelines.

👥
NEW

Team Rulesets

Shared rules committed to your repo. 6 starter templates (Python, React, Node, Go, Rust, Universal). Everyone starts with the same standards.

📚

Obsidian Dashboard

Every project syncs to your Obsidian vault. Specs, architecture, daily journals, quality verdicts — all browsable and linked.

🔌
NEW

Git Hooks

Pre-commit quality checks. Post-commit auto-learning. Context injection shows only the rules relevant to what you changed.

Up and running in 60 seconds

1

Install

pip install brainstormer — one dependency (PyYAML). Python 3.9+.

2

Init

cd your-project && brainstormer init — detects your stack, scaffolds docs, wires up all 4 skills.

3

Ship

Start coding. Rules accumulate. Walkthroughs appear. Run brainstormer status to see your project's health score.

Zero to BrainStormer

Never coded before? No problem. Here's exactly what to do.

Step 1: Get Python

Download Python 3.9+ from python.org. During install, check "Add Python to PATH".

Step 2: Open your terminal

Windows: Press Win + R, type cmd, press Enter. Mac: Open Spotlight, type Terminal. Linux: Ctrl + Alt + T.

Step 3: Install BrainStormer

Type pip install brainstormer and press Enter. That's it — one command.

Step 4: Initialize your project

Navigate to your project folder: cd path/to/your-project. Then run brainstormer init.

Step 5: See what happened

Run brainstormer status to see your project's health score. 10 new files appeared — they help AI understand your project.

$ brainstormer learn rule --from-diff HEAD~1

  Analyzing diff: a3c87cb -> HEAD (4 files changed)

  Proposed rule:
    Name:       Always tag agent source
    Do this:    Include source attribution in agent frontmatter
    Don't do:   Ship agents without source tags
    Confidence: high (3 hits, 0 misses)

  Rule saved to rules.md

$ brainstormer status

  Health: ||||||||||||||||.... 80%
  Rules:      12 (8 high confidence, 4 learning)
  Agents:     374 available, 5 recommended for Python
  Last sync:  2 minutes ago
  

Simple pricing

Start free. Upgrade when you need more.

Community

$0
Free forever
  • 3 projects
  • All 374 agents
  • All 4 skills (Kernel, Comprehension, Ideation, Quality)
  • Self-learning rules from diffs
  • One-way Obsidian sync (push)
  • Git hooks + context injection
  • Stack detection + migration
Get Started

What's next

BrainStormer's moat is knowledge accumulation. Each phase makes the next one more powerful.

Universal config sync Knowledge Graph API (MCP) Cross-tool session memory Multi-AI agent protocol 5-tier AI routing