Skip to content

Releases: Whatsonyourmind/oraclaw

OraClaw v1.0.1 — Decision Intelligence MCP Server

02 Apr 16:55

Choose a tag to compare

What's New

Billing & Monetization

  • Self-service signup: POST /api/v1/auth/signup — email to API key in one call
  • Stripe auto-provisioning of products, prices, and billing meters on boot
  • Webhook auto-setup for subscription lifecycle events
  • Usage _meta injection in all API responses (tier, calls remaining, upgrade hints)
  • Updated pricing: $0.005/call, $9/mo Starter, $49/mo Growth, $199/mo Scale

Distribution

  • Glama.ai listing: approved and live
  • x402 switched to Base mainnet for real USDC payments

MCP Server (v1.0.1)

  • 12 tools / 17 algorithms
  • Install: npx -y @oraclaw/mcp-server
  • All algorithms under 25ms, 14/17 under 1ms
  • 1,076 tests passing

API Endpoints

17 algorithm endpoints at https://oraclaw-api.onrender.com/api/v1/

  • Bandits, LP solver, Monte Carlo, Bayesian, graph analytics, CMA-ES, genetic algorithms, anomaly detection, forecasting, ensemble, scenario planning, pathfinding, convergence, calibration, risk analysis

OraClaw v2.3.0 -- 19 Algorithms, 12 MCP Tools, 14 npm Packages

30 Mar 22:27

Choose a tag to compare

OraClaw v2.3.0

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

Highlights

  • 19 algorithms covering optimization, simulation, prediction, detection, scoring, and planning
  • 12 MCP tools for seamless AI agent integration (Claude Code, Cursor, Windsurf, etc.)
  • 14 npm packages under @oraclaw/* -- install only what you need
  • Live REST API at https://oraclaw-api.onrender.com -- no signup, no API key for free tier
  • 1,072 tests passing with full benchmark suite
  • Sub-25ms latency on all endpoints (14 of 17 under 1ms)
  • USDC pay-per-call (x402) -- machine-to-machine payments on Base

Algorithm Suite

Category Algorithms
Optimization Multi-Armed Bandit, Contextual Bandit (LinUCB), CMA-ES, Genetic Algorithm
Simulation Monte Carlo, Scenario Planning
Solvers Constraint Solver (LP/MIP/QP via HiGHS), Schedule Optimizer
Analysis Decision Graph (PageRank, Louvain), Portfolio Risk (VaR/CVaR)
Prediction Bayesian Inference, Ensemble Model, Time Series Forecast
Detection Anomaly Detection (Z-Score, IQR)
Scoring Convergence, Calibration
Planning A* Pathfinding

npm Packages

npm install @oraclaw/bandit      # A/B testing with contextual bandits
npm install @oraclaw/solver      # LP/MIP resource allocation
npm install @oraclaw/cmaes       # Derivative-free optimization
npm install @oraclaw/graph       # PageRank, community detection
npm install @oraclaw/forecast    # ARIMA, Holt-Winters
npm install @oraclaw/anomaly     # Z-score, IQR detection
npm install @oraclaw/risk        # VaR, CVaR, stress testing
npm install @oraclaw/bayesian    # Bayesian inference
npm install @oraclaw/simulate    # Monte Carlo (6 distributions)
npm install @oraclaw/ensemble    # Multi-model consensus
npm install @oraclaw/calibrate   # Calibration scoring
npm install @oraclaw/evolve      # Genetic algorithm (Pareto)
npm install @oraclaw/pathfind    # A* pathfinding
npm install @oraclaw/decide      # Full decision pipeline

Quick Start

# No install needed -- hit the API directly
curl -X POST https://oraclaw-api.onrender.com/api/v1/optimize/bandit \
  -H 'Content-Type: application/json' \
  -d '{"arms": [{"id":"A","pulls":10,"totalReward":7},{"id":"B","pulls":2,"totalReward":1.8}], "algorithm":"ucb1"}'

Links

Architecture

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

Built on the OODA loop: Observe -> Orient -> Decide -> Act