OilGas Nanobot Swarm
ttracx/oilgas-nanobot-swarm
Multi-agent IA para engenharia upstream, midstream e downstream
Ver no GitHubSistema multi-agente hierárquico de IA para engenharia de petróleo e gás. Executa análises especializadas em perfuração, reservatórios, produção, dutos, petrofísica, controle de poço, segurança e economia. Funciona via API REST com Ollama Cloud e NVIDIA NIM, sem GPU local.
Instalação
npx convez-ai ttracx/oilgas-nanobot-swarm1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.
Métricas
Casos de uso
- Cálculos de perfuração e design de coluna
- Avaliação de reservatórios e curvas IPR/Vogel
- Análise de queda de pressão em dutos
- Modelagem petrofísica (porosidade, saturação)
- Controle de poço e kill sheets
- Análise econômica de projetos e AFE
Veredito

Sistema robusto e especializado para engenharia de óleo e gás com arquitetura de agentes bem estruturada. Pronto para produção, mas requer expertise técnica da indústria para máximo aproveitamento.
README
A tradução para o português deste README ainda está na fila.
OilGas Nanobot Swarm
Hierarchical AI Agent Swarm for Oil & Gas Engineering
Autonomous multi-agent engineering intelligence for upstream, midstream, and downstream operations
Live Demo · Quick Start · Architecture · Agent Teams · Tools · Deploy · API · Docs
Overview
OilGas Nanobot Swarm is an open-source, production-ready hierarchical multi-agent AI system purpose-built for oil and gas engineering. It runs on Vercel using Ollama Cloud and NVIDIA NIM as AI backends — no GPU required, no local model downloads.
Live Deployment
| URL | Status |
|---|---|
| oilgas-nanobot-swarm.vibecaas.app | Dashboard |
| oilgas-nanobot-swarm.vibecaas.app/health | Health check |
| oilgas-nanobot-swarm.vibecaas.app/docs | Swagger UI |
What It Does
| Domain | Capabilities |
|---|---|
| Drilling | ECD, kick tolerance, MAASP, well control kill sheets, casing design |
| Reservoir | IPR/Vogel curves, Darcy flow, formation evaluation, water saturation (Archie) |
| Production | Productivity index, decline curves, artificial lift selection |
| Pipeline | Pressure drop (Darcy-Weisbach), flow regime, line sizing |
| Petrophysics | Sonic porosity (Wyllie), shale volume (GR), Archie Sw, net pay |
| Well Control | Kill mud weight, MAASP, driller's method kill schedule |
| HSE & Regulatory | API standards, BSEE/BOEM, OSHA PSM (14 elements), EPA emissions |
| Economics | AFE preparation, NPV/IRR, break-even price, EUR sensitivity |
Architecture
┌─────────────────────────────────┐
│ FastAPI Gateway :8100 │
│ OpenAI-Compatible REST API │
└──────────────┬──────────────────┘
│
┌──────────────▼──────────────────┐
│ Queen Orchestrator (L0) │
│ Decomposes goals into L1 tasks │
└──┬──────┬──────┬──────┬─────────┘
│ │ │ │
┌───────────▼┐ ┌───▼──┐ ┌▼────┐ ┌▼──────────┐
│ Coder │ │Rsrch │ │Anlst│ │ Executor │ (L1)
└──────┬─────┘ └──┬───┘ └──┬──┘ └─────┬─────┘
│ │ │ │
┌───────▼──────────▼─────────▼───────────▼──────┐
│ L2 Sub-Agents (15 roles) │
└─────────────────────┬──────────────────────────┘
│
┌─────────────────────▼──────────────────────────┐
│ Engineering Tool Layer │
│ reservoir_pressure_calc · drilling_eng_calc │
│ production_eng_calc · pipeline_hydraulics │
│ well_control_calc · formation_eval_calc │
│ oilgas_regulatory_ref · web_search │
└────────────────────────────────────────────────┘
│
┌─────────────────────▼──────────────────────────┐
│ AI Backend Layer │
│ Primary: Ollama Cloud (ministral-3:8b) │
│ Fallback: NVIDIA NIM (llama-3.3-70b-instruct) │
└────────────────────────────────────────────────┘
Tier Structure
| Tier | Role | Count | Responsibility |
|---|---|---|---|
| L0 | Queen Orchestrator | 1 | Strategic decomposition and synthesis |
| L1 | Domain Leads | 6 | Coder, Researcher, Analyst, Validator, Executor, Architect |
| L2 | Sub-Agents | 15 | Narrow specialists (planner, writer, tester, searcher, etc.) |
Technology Stack
| Component | Technology |
|---|---|
| API Framework | FastAPI 0.115 |
| Primary AI | Ollama Cloud (ministral-3:8b, ~3-5s) |
| Fallback AI | NVIDIA NIM (meta/llama-3.3-70b-instruct) |
| State Store | Redis 7.4 (full-stack deployments) |
| Serialization | Pydantic v2 |
| HTTP Client | httpx 0.28 |
Quick Start
Option 1 — Vercel (Zero Setup, Live Now)
The app is already deployed. Grab the API key and start querying:
# Health check
curl https://oilgas-nanobot-swarm.vibecaas.app/health
# Engineering query (requires x-api-key header)
curl -X POST https://oilgas-nanobot-swarm.vibecaas.app/swarm/run \
-H "Content-Type: application/json" \
-H "x-api-key: YOUR_GATEWAY_KEY" \
-d '{"goal": "Calculate ECD at 10,000 ft TVD with 10.5 ppg mud and 320 psi APL"}'
Option 2 — GitHub Codespaces
# Inside Codespace — set env vars and start
export OLLAMA_API_KEY=your-ollama-key
export GATEWAY_API_KEY=your-gateway-key
python -m uvicorn nanobot.api.gateway:app --reload --port 8100
Option 3 — Local Docker Compose
git clone https://github.com/ttracx/oilgas-nanobot-swarm.git
cd oilgas-nanobot-swarm
# Configure
cp .env.example .env
# Edit .env — add OLLAMA_API_KEY and GATEWAY_API_KEY
# Start (includes Redis)
docker compose up -d
Option 4 — Local Python
git clone https://github.com/ttracx/oilgas-nanobot-swarm.git
cd oilgas-nanobot-swarm
pip install -e ".[dev]"
cp .env.example .env # configure OLLAMA_API_KEY
docker run -d -p 6379:6379 redis:7.4-alpine
python -m uvicorn nanobot.api.gateway:app --reload --port 8100
Configuration
# .env — copy from .env.example
# ── Primary AI Backend ─────────────────────────────────────────────────────────
# Ollama Cloud (recommended — fast, cost-effective)
OLLAMA_API_KEY=your-ollama-key
# Models available: ministral-3:8b, glm-5, kimi-k2:1t, qwen3-coder:480b, etc.
# ── Fallback AI Backend ────────────────────────────────────────────────────────
# NVIDIA NIM (high quality, robust fallback)
NVIDIA_API_KEY=nvapi-your-key
# Models: meta/llama-3.3-70b-instruct, moonshotai/kimi-k2-instruct-0905
# ── Optional: Anthropic Claude ────────────────────────────────────────────────
ANTHROPIC_API_KEY=sk-ant-your-key
ANTHROPIC_MODEL=claude-sonnet-4-20250514
# ── Gateway ────────────────────────────────────────────────────────────────────
GATEWAY_API_KEY=your-secure-api-key
# ── Redis State Store ──────────────────────────────────────────────────────────
REDIS_HOST=127.0.0.1
REDIS_PORT=6379
# ── Oil & Gas Features ─────────────────────────────────────────────────────────
ENABLE_OILGAS_TEAMS=true
Oil & Gas Agent Teams
| Team | Mode | Use Case |
|---|---|---|
well-engineering-review | Hierarchical | Casing design, ECD, kick tolerance, regulatory |
reservoir-analysis | Hierarchical | IPR curves, Sw, formation eval, reserve estimate |
drilling-ops-daily | Flat | Daily report, NPT tracking, cost per foot |
production-optimization | Hierarchical | AI lift, decline curves, waterflood analysis |
pipeline-integrity | Hierarchical | Corrosion, ILI review, MAOP verification |
hse-compliance-audit | Hierarchical | PSM 14-element audit, HAZOP, well integrity |
well-economics | Hierarchical | AFE, NPV/IRR, break-even, EUR sensitivity |
oilgas-field-briefing | Flat | Daily ops briefing, production summary, safety |
completions-design | Hierarchical | Frac design, stage spacing, proppant selection |
Example: Well Engineering Review
curl -X POST https://oilgas-nanobot-swarm.vibecaas.app/swarm/run \
-H "x-api-key: $GATEWAY_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"goal": "Review well X-1: TD 12,500 ft TVD, MW=10.5 ppg, shoe at 8,200 ft with FG=14.2 ppg. Check ECD, kick tolerance, and MAASP.",
"team": "well-engineering-review",
"mode": "hierarchical"
}'
Example: Daily Field Briefing
curl -X POST https://oilgas-nanobot-swarm.vibecaas.app/swarm/run \
-H "x-api-key: $GATEWAY_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"goal": "Generate daily field briefing. Production: 8,200 BOPD, 14.5 MMSCFD gas, 22,000 BWPD. 2 wells shut-in for ESP replacement.",
"team": "oilgas-field-briefing"
}'
Engineering Tools
| Tool | Calculates |
|---|---|
reservoir_pressure_calc | Hydrostatic gradient, BHP, pore pressure, fracture gradient |
drilling_engineering_calc | ECD, kick tolerance, mud weight window, surge/swab |
production_engineering_calc | PI, Vogel IPR, Darcy flow, artificial lift selection |
pipeline_hydraulics_calc | Pressure drop (D-W), flow regime, line sizing |
well_control_calc | MAASP, kill mud weight, driller's method |
formation_evaluation_calc | Archie Sw, Wyllie sonic porosity, GR shale volume |
oilgas_regulatory_reference | API standards, BSEE, OSHA PSM, EPA emissions |
API Reference
| Method | Path | Auth | Description |
|---|---|---|---|
GET | /health | — | System health check |
GET | / | — | Web dashboard |
GET | /docs | — | Interactive Swagger UI |
POST | /swarm/run | API key | Dispatch an engineering task |
POST | /v1/chat/completions | — | OpenAI-compatible chat |
GET | /v1/models | — | List available models |
GET | /swarm/health | — | Swarm status |
GET | /swarm/topology | — | Agent hierarchy |
/swarm/run Request
{
"goal": "Your engineering question or task",
"mode": "hierarchical",
"team": "optional-team-name",
"metadata": {}
}
/swarm/run Response
{
"success": true,
"session_id": "vercel-1234567890",
"goal": "...",
"final_answer": "Full engineering analysis with calculations...",
"subtask_count": 1,
"duration_seconds": 7.6,
"mode": "vercel-serverless",
"model": "ministral-3:8b"
}
Deployment
Deploy to Vercel
npm i -g vercel
vercel login
vercel --prod
Required environment variables in Vercel dashboard:
| Variable | Description |
|---|---|
OLLAMA_API_KEY | Ollama Cloud API key (primary AI) |
NVIDIA_API_KEY | NVIDIA NIM API key (fallback AI) |
GATEWAY_API_KEY | Auth key for /swarm/run |
Deploy to Render
The render.yaml provisions: web service + managed Redis + 5 GB disk.
Deploy to Railway
npm i -g @railway/cli && railway login && railway up
Open in GitHub Codespaces
Scheduled Workflows
Configure recurring agent team runs in workspace/schedules.json:
[
{
"name": "daily-field-briefing",
"team": "oilgas-field-briefing",
"goal": "Generate the 7am daily field operations briefing.",
"schedule": "07:00",
"enabled": true
},
{
"name": "weekly-production-review",
"team": "production-optimization",
"goal": "Analyze production performance for all wells over the past 7 days.",
"schedule": "monday 08:00",
"enabled": true
}
]
Start the scheduler: python -m nanobot.scheduler.scheduler
Project Structure
oilgas-nanobot-swarm/
├── api/
│ ├── index.py # Vercel serverless entry (Ollama + NVIDIA NIM)
│ └── requirements.txt # Slim deps for Vercel function
├── nanobot/
│ ├── core/ # Agent hierarchy (Queen L0, L1 leads, L2 sub-agents)
│ ├── tools/
│ │ ├── oilgas_tools.py ← 7 engineering calculators
│ │ ├── web_search.py
│ │ └── code_runner.py
│ ├── teams/
│ │ └── oilgas_teams.py ← 9 pre-configured O&G agent teams
│ ├── scheduler/ # Background job scheduler
│ ├── knowledge/ # Markdown vault (Obsidian-compatible)
│ ├── state/ # Redis persistence
│ ├── integrations/ # External services (MS Graph, OpenClaw)
│ ├── api/ # FastAPI gateway (full-stack mode)
│ └── static/
│ └── index.html # NeuralQuantum-branded web dashboard
├── docs/ # Comprehensive documentation
│ ├── ARCHITECTURE.md
│ ├── API.md
│ ├── AGENTS.md
│ ├── DEPLOYMENT.md
│ ├── ENGINEERING_TOOLS.md
│ ├── AGENT_TEAMS.md
│ └── DEVELOPMENT.md
├── scripts/ # Setup and launch scripts
├── .devcontainer/ # GitHub Codespaces config
├── docker-compose.yml # Local dev stack (gateway + Redis)
├── Dockerfile # Production container
├── render.yaml # Render deployment
├── railway.toml # Railway deployment
├── vercel.json # Vercel deployment
└── pyproject.toml
Safety Disclaimer
⚠️ ENGINEERING DISCLAIMER: This system is a decision-support tool for qualified petroleum engineers. All calculations, regulatory references, and recommendations must be verified by licensed engineers before use in operational decisions.
Pro Edition
⭐ Pro Edition Coming Soon — Full hierarchical swarm, Redis memory, knowledge vault, background scheduler, MS 365 integration, and priority support.
Contact: info@neuralquantum.ai
Contributing
git checkout -b feature/my-tool
# implement, add tests
pytest tests/ -v
git push && open PR
License
MIT License — see LICENSE.
Powered by VibeCaaS.com · a division of NeuralQuantum.ai LLC
© 2026 OilGas Nanobot Swarm powered by VibeCaaS.com a division of NeuralQuantum.ai LLC. All rights reserved.