AI agents that think like military strategists.

# Option 1: Git submodule
git submodule add https://github.com/drwestman/ooda-subagents.git .claude

# Option 2: Copy agents
git clone https://github.com/drwestman/ooda-subagents.git
cp -r ooda-subagents/agents <your-project>/.claude/

Observe → Orient → Decide → Act. A decision-making loop that turns chaos into clarity.

┌─────────────┐
│   OBSERVE   │ ← Gather info
└──────┬──────┘
       ▼
┌─────────────┐
│   ORIENT    │ ← Analyze patterns
└──────┬──────┘
       ▼
┌─────────────┐
│   DECIDE    │ ← Pick best path
└──────┬──────┘
       ▼
┌─────────────┐
│     ACT     │ ← Execute
└─────────────┘

Gathers info: reads files, searches code, finds patterns.

Makes sense of data: spots connections, analyzes context.

Weighs options: evaluates trade-offs, recommends solutions.

Executes: writes code, runs tests, ships features.

Debugging an auth issue:

1. Observe: Scans error logs and recent commits
2. Orient: Identifies race condition pattern
3. Decide: Recommends mutex locking strategy
4. Act: Implements fix and verifies

• Systematic problem-solving
• Faster delivery through structured thinking
• Nothing falls through cracks

We help early-stage startups build AI-first products that ship fast. OODA agents are how we deliver speed AND quality.

What we do:

• Build AI products that ship (not demos)
• Set up scalable dev processes
• Turn vision into working code

Got an AI product idea? Let's build → astromvp.com