{"id":"7a3c099f-8931-42e8-9a51-1d69c57095e1","slug":"archonai","kind":"mcp","name":"ArchonAI","summary":"ARCHON — Your Frontier AI Architect. 6 autonomous specialist agents audit your codebase in parallel, combining RAG + live web research to deliver complete Architecture Review Packages with ADRs, IaC s","owner":"VenkataAnilKumar","repoUrl":"https://github.com/VenkataAnilKumar/ArchonAI","license":"MIT","licenseTier":"permissive","trust":"community","stars":1,"match":null,"fewStars":true,"format":"mcp","createdAt":"2026-03-31T22:41:22.000Z","addedAt":"2026-10-21T08:37:07.464Z","logoUrl":"https://avatars.githubusercontent.com/u/25908037?u=17538b595aa6cc626191ed955737cf316b3f57b5&v=4","linkOnly":false,"collection":null,"skillPath":null,"categories":["search","developer-tools"],"tags":["ai-agents","artificial-intelligence","code","llms","rag","agentic-ai","architecture","autnomous"],"runtime":"python","local":true,"cloudCapable":false,"cloudCapability":"unknown","cloudReadyOnPlatform":false,"needsApiKey":false,"reviewed":false,"syncedAt":"2026-10-23T03:29:14.707Z","sample":false,"url":"https://conduits.dev/tools/archonai","language":"Python","transports":["stdio"],"install":{"pypi":"archon"},"requirements":{"env":[],"upstreamMayCharge":false},"tools":null,"readme":"
\n\n# ARCHON\n\n### Your Frontier AI Architect. From idea to infrastructure.\n\n[![Python 3.11+](https://img.shields.io/badge/Python-3.11+-3776AB?style=for-the-badge&logo=python&logoColor=white)](https://python.org)\n[![Next.js 15](https://img.shields.io/badge/Next.js-15-000000?style=for-the-badge&logo=next.js&logoColor=white)](https://nextjs.org)\n[![FastAPI](https://img.shields.io/badge/FastAPI-0.115+-009688?style=for-the-badge&logo=fastapi&logoColor=white)](https://fastapi.tiangolo.com)\n[![Claude Opus 4](https://img.shields.io/badge/Claude-Opus_4-CC785C?style=for-the-badge&logo=anthropic&logoColor=white)](https://anthropic.com)\n[![Tests](https://img.shields.io/badge/Tests-531_passing-brightgreen?style=for-the-badge&logo=pytest&logoColor=white)](src/tests/)\n[![Coverage](https://img.shields.io/badge/Coverage-89%25-brightgreen?style=for-the-badge)](src/tests/)\n[![License](https://img.shields.io/badge/License-MIT-green?style=for-the-badge)](LICENSE)\n\n**6 specialist AI architects. 15 modes. Parallel execution. Cited findings. One package.**\n\n[Getting Started](#-getting-started) · [Features](#-features) · [Modes](#-the-15-modes) · [Architecture](#-architecture) · [Distribution](#-distribution) · [Pricing](#-pricing)\n\n
\n\n---\n\n![ARCHON Demo](./assets/demo.svg)\n\n---\n\n## The Problem\n\nHiring a principal architect costs **$200–350k/year**. External consultants charge **$5–15k/engagement** and take weeks. Most startups skip architecture reviews entirely — and the average architectural mistake costs **$50–500k to fix** post-production.\n\n## The Solution\n\nARCHON is an **autonomous AI architecture co-pilot** that delivers a complete, cited Architecture Review Package in under 60 minutes.\n\n- **Frontier Agent Autonomy** — 6 specialist architects run in parallel for up to 60 minutes\n- **Perplexity-Style Research** — live web search (Tavily + Exa) with every finding cited and scored\n- **Codebase Intelligence** — RAG over your actual code, not generic advice\n- **Human-in-the-Loop** — 4 checkpoints, 3 control modes (Autopilot / Balanced / Supervised)\n- **15 Modes** — from architecture review to PR gating, cost optimisation, and incident response\n- **MCP Server** — expose ARCHON as an MCP tool for Claude Desktop, VS Code, and custom agents\n\n---\n\n## Features\n\n| Feature | Description |\n|---|---|\n| **6 Specialist Agents** | Software, Cloud, Security, Data, Integration, and AI architects |\n| **Parallel Execution** | All 6 agents run simultaneously via `asyncio.gather()` |\n| **15 Modes** | Review, Design, Migration, Compliance, PR Reviewer, Cost Optimiser, and 9 more |\n| **Full Citations** | Every finding backed by code evidence or a live web source URL |\n| **Confidence Scoring** | Per-finding and per-section confidence with methodology breakdown |\n| **Mermaid Diagrams** | Auto-generated C4, data flow, integration, and deployment diagrams |\n| **Architecture Memory** | Persistent decision history and architecture snapshot diffing |\n| **Health Score** | 0–100 architecture health score with SVG dashboard |\n| **MCP Server** | 5 MCP tools: review, findings, ask, health score, ADRs |\n| **HITL Checkpoints** | Pause, review, and redirect agents at 4 decision points |\n| **Export Formats** | Markdown, ZIP package, HTML, PDF, GitHub comment, Slack message |\n| **Multi-Input** | GitHub URL, ZIP, PDF, Figma, website, IaC, OpenAPI spec, voice, image |\n| **VS Code Extension** | Run reviews and browse findings directly from your editor |\n| **GitHub App** | Auto-review PRs on open/sync — posts structured comment within 2 min |\n| **CLI Package** | `pip install archon-cli` — login, review, stream, export |\n| **Zero-Trust Security** | Clerk JWT auth, per-tenant isolation, secrets never logged |\n| **Stripe Billing** | 3 subscription tiers + event-based pricing |\n\n---\n\n## The 6 Agents\n\n```\n+-------------------------------------------------------------+\n| SUPERVISOR |\n| Fan-out → Deduplicate → Merge |\n+----------+----------+----------+----------+----------+------+\n| Software | Cloud | Security | Data | Integr. | AI |\n| Architect| Architect| Architect| Architect| Architect| Arch |\n| | | | | | |\n| Patterns | AWS/GCP | Zero | Schema | EDA | RAG |\n| NFRs | Azure | Trust | Govern. | APIs | ML |\n| Tech Debt| IaC | IAM | Pipelines| Service | Agent|\n| ADRs | FinOps | Comply | PII | Mesh | Ops |\n+----------+----------+----------+----------+----------+------+\n```\n\n---\n\n## The 15 Modes\n\n
\nClick to expand all modes\n\n| # | Mode | Category | Description |\n|---|---|---|---|\n| 1 | **Review** | Lifecycle | Audit existing codebase — the core mode |\n| 2 | **Design** | Lifecycle | New product architecture from a brief |\n| 3 | **Migration Planner** | Lifecycle | Modernisation and platform migration |\n| 4 | **Compliance Auditor** | High Urgency | SOC2 / HIPAA / GDPR / PCI-DSS audit |\n| 5 | **Due Diligence** | High Urgency | Investor / M&A technical package |\n| 6 | **Incident Responder** | High Urgency | P0/P1 outage architecture triage |\n| 7 | **Cost Optimiser** | Continuous | Cloud spend reduction with savings matrix |\n| 8 | **PR Reviewer** | Continuous | Architecture review on pull requests |\n| 9 | **Scaling Advisor** | Continuous | Capacity and growth planning from APM data |\n| 10 | **Drift Monitor** | Continuous | Weekly architecture health check |\n| 11 | **Feature Feasibility** | Decision Support | Build / buy / defer verdict for a feature brief |\n| 12 | **Vendor Evaluator** | Decision Support | Database / cloud / tool TCO + lock-in matrix |\n| 13 | **Onboarding Accelerator** | Decision Support | Annotated system map + learning path for new hires |\n| 14 | **Sunset Planner** | Decision Support | Service shutdown sequence + GDPR checklist |\n| 15 | **Idea Mode** | Design | Natural language → complete architecture in minutes |\n\n
\n\n---\n\n## Getting Started\n\n### Prerequisites\n\n- Python 3.11+\n- Docker and Docker Compose\n- Node.js 18+ (for frontend)\n- API keys: Anthropic, Tavily, Exa\n\n### 1. Clone and Install\n\n```bash\ngit clone https://github.com/VenkataAnilKumar/ArchonAI.git\ncd ArchonAI\n\n# Install Python dependencies\npip install uv\nuv pip install -e \".[dev]\"\n\n# Install frontend dependencies\ncd web && npm install && cd ..\n```\n\n### 2. Configure Environment\n\n```bash\ncp .env.example .env\n# Edit .env with your API keys:\n# ANTHROPIC_API_KEY=sk-ant-...\n# TAVILY_API_KEY=tvly-...\n# EXA_API_KEY=exa-...\n# DATABASE_URL=postgresql://...\n# REDIS_URL=redis://...\n```\n\n### 3. Start Infrastructure\n\n```bash\ndocker compose up -d # PostgreSQL (pgvector) + Redis\n```\n\n### 4. Run Database Migrations\n\n```bash\ncd src/db && alembic upgrade head && cd ../..\n```\n\n### 5. Run\n\n**CLI (quickest)**\n\n```bash\n# Review an existing codebase\npython main.py --repo https://github.com/user/repo --mode review\n\n# Design from a brief\npython main.py --brief \"SaaS video platform, 10k users, $2k/month budget\" --mode design\n\n# Check feature feasibility\npython main.py --brief \"Add multi-region failover in 6 weeks\" --mode feature_feasibility\n\n# With Human-in-the-Loop\npython main.py --repo https://github.com/user/repo --mode review --hitl balanced\n\n# See all 15 modes\npython main.py --help\n```\n\n**Web App**\n\n```bash\n# Terminal 1: API\nuvicorn src.api.main:app --reload --port 8000\n\n# Terminal 2: Frontend\ncd web && npm run dev\n\n# Open http://localhost:3000\n```\n\n**CLI Package**\n\n```bash\npip install archon-cli\n\narchon login # authenticate with your ARCHON account\narchon review --repo https://github.com/user/repo\narchon status # check account + active reviews\n```\n\n---\n\n## Distribution\n\nARCHON ships across four surfaces — all code-complete:\n\n### VS Code Extension\nInstall from VS Code Marketplace (coming soon) or build locally:\n```bash\ncd vscode-extension && npm install && npm run compile\n```\n- `ARCHON: Configure API Key` — stores key in VS Code SecretStorage\n- `ARCHON: Review Workspace` — trigger review on open project\n- Findings sidebar, inline decorations, live SSE progress panel\n\n### GitHub App\nInstall from GitHub Marketplace (coming soon). Once installed:\n- Opens a PR → ARCHON posts a structured architecture review comment within 2 min\n- Per-repo config via `.archon.yml` (branches, ignore patterns, file cap)\n- Retry logic, partial-failure comments, rate limiting built in\n\n### CLI Package (`archon-cli`)\n```bash\npip install archon-cli # coming soon on PyPI\n```\nFull command set: `login`, `logout`, `status`, `review`, `design`, `export`\nRich live SSE display — watch all 6 agents run in real time in your terminal.\n\n### MCP Server\nExpose ARCHON as an MCP server for Claude Desktop or any MCP-compatible client:\n```json\n{\n \"mcpServers\": {\n \"archon\": {\n \"command\": \"python\",\n \"args\": [\"-m\", \"archon.mcp.server\"],\n \"env\": { \"ARCHON_API_KEY\": \"your-key\" }\n }\n }\n}\n```\nAvailable tools: `review_repo`, `get_findings`, `ask_architecture`, `get_health_score`, `get_adrs`\n\n---\n\n## Architecture\n\n```\nArchon/\n├── main.py # CLI entry point (15 modes)\n├── src/\n│ ├── archon/ # Core domain (hexagonal architecture)\n│ │ ├── core/\n│ │ │ ├── models/ # Pydantic domain models\n│ │ │ └── ports/ # Abstract interfaces (ports)\n│ │ ├── agents/ # 6 specialist agents + base class\n│ │ ├── engine/ # Supervisor, Runner, HITL, 15 mode configs\n│ │ ├── engine/modes/ # Per-mode focus, inputs, output sections\n│ │ ├── rag/ # Chunker, Indexer, Retriever\n│ │ ├── input/ # Multi-format parsers (PDF, APM, cost CSV, IaC)\n│ │ ├── output/ # Formatters, ZIP builder, diagrams, share\n│ │ ├── mcp/ # MCP server (5 tools)\n│ │ ├── workers/ # Drift worker (cron-triggered)\n│ │ └── infrastructure/ # Claude, Tavily, Exa, pgvector, GitHub, Redis\n│ ├── api/ # FastAPI (9 route groups, SSE streaming)\n│ ├── db/ # SQLAlchemy models + Alembic migrations\n│ └── tests/ # 513 pytest tests, 89% coverage\n├── web/ # Next.js 15 + Clerk + Stripe + shadcn/ui\n├── vscode-extension/ # VS Code extension (TypeScript)\n├── cli-package/ # archon-cli PyPI package\n├── github-app/ # GitHub App for PR auto-review\n└── docs/ # PRDs, ADRs, architecture, runbooks\n```\n\n### Clean Architecture (Hexagonal)\n\n```\n+--------------------------------------------------+\n| Delivery Layer |\n| CLI · FastAPI · Next.js · VS Code · GitHub App |\n+--------------------------------------------------+\n| Application Layer |\n| Supervisor · Runner · HITL · RAG · Output |\n+--------------------------------------------------+\n| Domain Layer (zero external dependencies) |\n| core/models (Pydantic) · core/ports (ABCs) |\n+--------------------------------------------------+\n| Infrastructure Layer |\n| Claude · Tavily · Exa · pgvector · GitHub |\n+--------------------------------------------------+\n Dependency direction: inward only\n```\n\n---\n\n## Output Package\n\nEvery ARCHON review produces a complete, downloadable package:\n\n```\narchon-review-myrepo-20260401/\n├── README.md # Executive summary + navigation guide\n├── findings/ # One .md per agent domain (6 files)\n├── adrs/ # Architecture Decision Records\n├── terraform/ # IaC skeletons with comments\n├── diagrams/ # Mermaid .mmd files (C4, data flow, deployment)\n├── risk-register.md # All findings sorted by severity + confidence\n└── citations.md # All sources consolidated and deduplicated\n```\n\n---\n\n## Testing\n\n```bash\n# Full suite\npytest src/tests/ cli-package/tests/ github-app/tests/ -q\n\n# With coverage\npytest src/tests/ --cov=src --cov-report=term-missing\n\n# Single suite\npytest src/tests/ -v --tb=short\n```\n\n**531 tests · 89% coverage · 0 warnings**\n\n---\n\n## Pricing\n\n| Plan | Price | Reviews | Modes |\n|---|---|---|---|\n| **Starter** | $49/month | 3 on-demand | Core 6 modes |\n| **Pro** | $199/month | Unlimited | All 15 modes + PR Reviewer |\n| **Team** | $499/month | Unlimited | All modes + Drift Monitor + 5 seats |\n\n**Event-based (one-shot):**\n\n| Mode | Price |\n|---|---|\n| Due Diligence Responder | $999/run |\n| Compliance Auditor | $499/run |\n| Migration Planner | $499/run |\n| Incident Responder | $299/run |\n\n---\n\n## Documentation\n\n| Document | Description |\n|---|---|\n| [Product Plan](docs/PRODUCT_PLAN.md) | Full product vision, pricing, GTM |\n| [Architecture](docs/ARCHITECTURE.md) | System design + hexagonal architecture |\n| [PRDs](docs/prds/) | 15 mode PRDs with user stories + NFRs |\n| [ADRs](docs/architecture/) | Architecture Decision Records |\n| [Getting Started](docs/development/getting-started.md) | Developer setup guide |\n| [Security Model](docs/security/security-model.md) | Data handling + tenant isolation |\n| [Runbooks](docs/runbooks/) | DB migrations, incidents, scaling |\n\n---\n\n## Contributing\n\n1. **Fork** the repository\n2. **Create** a feature branch: `git checkout -b feature/amazing-feature`\n3. **Follow** the coding conventions in [`.claude/rules/python.md`](.claude/rules/python.md)\n4. **Write tests** — maintain ≥ 89% coverage\n5. **Run** `ruff check .` and `pytest` before committing\n6. **Submit** a pull request\n\n### Code Standards\n\n- Python 3.11+ with type hints on all public functions\n- `ruff` for linting and formatting\n- Pydantic v2 models for all data contracts\n- Async I/O everywhere (`httpx`, not `requests`)\n- Google-style docstrings on public APIs\n- `MagicMock()` for containers, `AsyncMock()` for async methods in tests\n\n---\n\n## License\n\nMIT License. See [LICENSE](LICENSE) for details.\n\n---\n\n## Contact\n\n- **Author:** Venkata Anil Kumar\n- **GitHub:** [@VenkataAnilKumar](https://github.com/VenkataAnilKumar)\n- **Project:** [github.com/VenkataAnilKumar/ArchonAI](https://github.com/VenkataAnilKumar/ArchonAI)\n\n---\n\n
\n\n**Built by architects, for architects.**\n\n*ARCHON — because every startup deserves a principal architect.*\n\n
\n"}