OraClaw

Decision intelligence for AI agents. 19 algorithms, 12 MCP tools, sub-25ms. Zero LLM cost.

OraClaw gives your AI agent real decision-making capabilities -- bandits, solvers, forecasters, risk models -- without burning tokens on reasoning. Every tool runs deterministically, returns structured JSON, and costs nothing to compute.

Try It Now

The API is live. No signup, no API key, no SDK install. Just curl:

curl -X POST https://oraclaw-api.onrender.com/api/v1/optimize/bandit \
  -H 'Content-Type: application/json' \
  -d '{
    "arms": [
      {"id": "A", "name": "Option A", "pulls": 10, "totalReward": 7},
      {"id": "B", "name": "Option B", "pulls": 10, "totalReward": 5},
      {"id": "C", "name": "Option C", "pulls": 2, "totalReward": 1.8}
    ],
    "algorithm": "ucb1"
  }'

Response (<1ms):

{
  "selected": { "id": "C", "name": "Option C" },
  "score": 1.876,
  "algorithm": "ucb1",
  "exploitation": 0.9,
  "exploration": 0.976,
  "regret": 0.1
}

The bandit correctly identifies Option C as under-explored and selects it for testing -- balancing exploration and exploitation without you writing any statistics code.

Quick Start

Option 1: REST API (no install)

Hit any of 17 endpoints directly. Free tier (100 calls/day) needs no auth.

# Bayesian inference
curl -X POST https://oraclaw-api.onrender.com/api/v1/predict/bayesian \
  -H 'Content-Type: application/json' \
  -d '{"prior": 0.3, "evidence": [{"factor": "positive_test", "weight": 0.9, "value": 0.05}]}'

# Monte Carlo simulation
curl -X POST https://oraclaw-api.onrender.com/api/v1/simulate/montecarlo \
  -H 'Content-Type: application/json' \
  -d '{"simulations": 1000, "distribution": "normal", "params": {"mean": 100, "stddev": 15}}'

# Anomaly detection
curl -X POST https://oraclaw-api.onrender.com/api/v1/detect/anomaly \
  -H 'Content-Type: application/json' \
  -d '{"data": [10, 12, 11, 13, 50, 12, 11, 10], "method": "zscore", "threshold": 2.0}'

Option 2: MCP Server (for AI agents)

Add to your Claude Code mcp.json or claude_desktop_config.json:

{
  "mcpServers": {
    "oraclaw": {
      "command": "npx",
      "args": ["tsx", "/path/to/oraclaw/packages/mcp-server/src/index.ts"]
    }
  }
}

Then ask your agent:

"Which email subject line should I use? Here are results from the last 3 variants..."

The agent calls optimize_bandit and gets a statistically grounded selection in 0.01ms.

Option 3: npm SDK

npm install @oraclaw/bandit

import { OraBandit } from '@oraclaw/bandit';

const client = new OraBandit({ baseUrl: 'https://oraclaw-api.onrender.com' });
const result = await client.optimize({
  arms: [
    { id: 'A', name: 'Short Subject', pulls: 500, totalReward: 175 },
    { id: 'B', name: 'Long Subject', pulls: 300, totalReward: 126 },
  ],
  algorithm: 'ucb1',
});

14 SDK packages available: @oraclaw/bandit, @oraclaw/solver, @oraclaw/simulate, @oraclaw/risk, @oraclaw/forecast, @oraclaw/anomaly, @oraclaw/graph, @oraclaw/bayesian, @oraclaw/ensemble, @oraclaw/calibrate, @oraclaw/evolve, @oraclaw/pathfind, @oraclaw/cmaes, @oraclaw/decide

Algorithms (19)

14 of 17 endpoints respond in under 1ms. All under 25ms.

MCP Tools (12)

Pricing

USDC pay-per-call (x402): AI agents pay $0.01-$0.15 per call autonomously with USDC on Base. No subscription, no API key. Machine-to-machine payments.

Performance

All benchmarks measured over 50 iterations on a single core. No GPU, no external services.

Products (14 npm packages)

Architecture

Built on the OODA loop framework:

OBSERVE  -->  Signal detection, anomaly scanning, data ingestion
ORIENT   -->  Context building, graph analysis, environment mapping
DECIDE   -->  Option generation, Monte Carlo simulation, Bayesian inference
ACT      -->  Execution planning, pathfinding, real-time adjustment

Stack: Fastify 5 + TypeScript (strict) | Expo 55 + React Native 0.83 | PostgreSQL (Supabase) | Turborepo monorepo

Project Structure

oraclaw/
  mission-control/
    apps/
      api/           # Fastify backend (17 endpoints, 19 algorithms)
      mobile/        # Expo React Native app
    packages/
      mcp-server/    # MCP server (12 tools for AI agents)
      sdk/           # 14 npm packages (@oraclaw/*)
      clawhub-skills/# 14 ClawHub skills for agent marketplaces
      shared-types/  # TypeScript interfaces
      schemas/       # Zod validation
  web/               # Next.js 15 dashboard + docs
  launch/            # GTM materials

Development

cd mission-control

# Start dev server
npm run dev

# Run tests (1,072 passing)
npm run test

# Run benchmarks
npx tsx scripts/benchmark-all.ts

# Type check
cd apps/api && npx tsc --noEmit

Links

• Live API: https://oraclaw-api.onrender.com
• npm: https://www.npmjs.com/org/oraclaw
• GitHub: https://github.com/Whatsonyourmind/oracle

License

MIT