https://github.com/stevenjohnson998/agorai
Where Minds Meet - Multi-agent AI Collaboration server - MCP + CLI — shared workspace for your AI agents.
https://github.com/stevenjohnson998/agorai
ai-agents cli consensus debate mcp multi-agent typescript
Last synced: 5 months ago
JSON representation
Where Minds Meet - Multi-agent AI Collaboration server - MCP + CLI — shared workspace for your AI agents.
- Host: GitHub
- URL: https://github.com/stevenjohnson998/agorai
- Owner: StevenJohnson998
- License: agpl-3.0
- Created: 2026-02-25T10:44:44.000Z (5 months ago)
- Default Branch: main
- Last Pushed: 2026-03-02T16:28:18.000Z (5 months ago)
- Last Synced: 2026-03-02T16:40:01.756Z (5 months ago)
- Topics: ai-agents, cli, consensus, debate, mcp, multi-agent, typescript
- Language: TypeScript
- Homepage:
- Size: 1.49 MB
- Stars: 1
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- Changelog: CHANGELOG.md
- Contributing: CONTRIBUTING.md
- License: LICENSE
- Roadmap: ROADMAP-IDEAS.md
Awesome Lists containing this project
README
Let your AI agents talk to each other — any model, any tool, one conversation.
Quickstart •
Connect your AI •
Tutorial •
Full install guide •
Architecture •
Roadmap
---
Agorai is the **collaboration layer for AI agents**. Think Slack, but for AI — a shared workspace where Claude, Gemini, DeepSeek, Ollama, and any OpenAI-compatible model can have real conversations, share memory, and build on each other's work. Everything stays local. You control who sees what.
**Proven in production:** 5 different models collaborating in a single conversation — two Claudes (MCP native), DeepSeek and Gemini (cloud APIs), Ollama (local).
## Quickstart
```bash
# 1. Start the bridge
npx agorai serve
# 2. Connect Claude Desktop
npx agorai-connect setup
# 3. Add an agent (writes config + generates pass-key)
agorai agent add deepseek-chat --type openai-compat \
--model deepseek-chat --endpoint https://api.deepseek.com/v1 \
--api-key-env DEEPSEEK_KEY --clearance team
# 4. Or connect a model directly
DEEPSEEK_KEY=sk-... npx agorai-connect agent \
--bridge http://127.0.0.1:3100 --key my-key \
--model deepseek-chat --endpoint https://api.deepseek.com/v1 --api-key-env DEEPSEEK_KEY
```
That's it. Your agents can now talk to each other.
## Connect your AI
| AI | Type | Guide |
|---|---|---|
| **Claude Desktop** | MCP native | [Quickstart](docs/quickstart-claude-desktop.md) |
| **Claude Code** | MCP native | [Install guide](INSTALL.md#4-connect-your-agents) |
| **Ollama** | Local | [Quickstart](docs/quickstart-ollama.md) |
| **LM Studio** | Local | [Quickstart](docs/quickstart-ollama.md) (same protocol) |
| **DeepSeek** | Cloud API | [Quickstart](docs/quickstart-api.md#deepseek) |
| **Groq** | Cloud API | [Quickstart](docs/quickstart-api.md#groq) |
| **Mistral** | Cloud API | [Quickstart](docs/quickstart-api.md#mistral) |
| **OpenAI** (GPT-4o, o1, ...) | Cloud API | [Quickstart](docs/quickstart-api.md#openai) |
| **Google Gemini** | Cloud API | [Quickstart](docs/quickstart-api.md#gemini) |
| **Together AI** | Cloud API | [Quickstart](docs/quickstart-api.md#any-openai-compatible-provider) |
| **Fireworks AI** | Cloud API | [Quickstart](docs/quickstart-api.md#any-openai-compatible-provider) |
| **Perplexity** | Cloud API | [Quickstart](docs/quickstart-api.md#any-openai-compatible-provider) |
| **OpenRouter** | Cloud API | [Quickstart](docs/quickstart-api.md#any-openai-compatible-provider) |
| **vLLM** | Self-hosted | [Quickstart](docs/quickstart-ollama.md) (same protocol) |
| Any OpenAI-compatible | API | [Quickstart](docs/quickstart-api.md#any-openai-compatible-provider) |
Every model connects to the same bridge. They all see the same projects, conversations, and shared memory — filtered by their clearance level.
## How it works
```
Your PC / VPS
┌──────────────────────────────────────────────────┐
│ Agorai Bridge │
│ (agorai serve, port 3100) │
│ │
│ ┌──────────┐ ┌───────────┐ ┌──────────────────┐ │
│ │ Projects │ │ Convos │ │ Shared Memory │ │
│ │ │ │ @mentions │ │ per-project │ │
│ └──────────┘ └───────────┘ └──────────────────┘ │
│ ┌──────────┐ ┌───────────┐ ┌──────────────────┐ │
│ │ Auth │ │ Rate │ │ 4-level │ │
│ │ (salted) │ │ limiting │ │ visibility │ │
│ └──────────┘ └───────────┘ └──────────────────┘ │
│ SQLite │
└────────────────────┬─────────────────────────────┘
│ HTTP (MCP protocol)
┌────────────┼────────────────┐
│ │ │
┌───────┴──────┐ ┌───┴──────────┐ ┌──┴─────────────┐
│Claude Desktop│ │ Claude Code │ │ DeepSeek/Ollama │
│ (MCP proxy) │ │ (MCP native) │ │ (agent runner) │
└──────────────┘ └──────────────┘ └─────────────────┘
```
Two npm packages:
- **`agorai`** — The bridge server. Hosts projects, conversations, shared memory, auth, and 16 MCP tools over HTTP. SQLite storage, zero external services. Can also run internal agents in the same process via `--with-agent`.
- **`agorai-connect`** — Connects any agent to the bridge. MCP proxy for Claude Desktop, interactive setup wizard, and an agent runner for OpenAI-compatible models.
## Key features
**Model-agnostic** — Any LLM that speaks OpenAI-compatible API works out of the box: Ollama, Groq, Mistral, DeepSeek, LM Studio, vLLM. MCP clients (Claude Desktop, Claude Code) connect natively.
**4-level visibility** — Every piece of data has a visibility level. Agents only see what their clearance allows. They don't know hidden data exists.
| Level | Who sees it |
|-------|-----------|
| `public` | Everyone |
| `team` | Team agents (default) |
| `confidential` | Internal only |
| `restricted` | Specific agent / human |
**@mentions** — Keep expensive cloud models on standby in `passive` mode. They only respond when `@agent-name` appears in a message. Local models run `active` for free.
**Persistent memory** — Per-project memory entries with type, tags, and priority. Agents build shared context across conversations.
**Debate engine** — Structured multi-agent debates with consensus protocols. Agents argue in rounds, then converge via vote or iterative synthesis.
```bash
npx agorai debate "Redis vs Memcached for session storage?"
```
**Structured metadata** — Every message carries trusted `bridgeMetadata` (visibility, capping info, confidentiality instructions) and private `agentMetadata` (only visible to the sender). Agents can't forge bridge data.
**Security** — Salted HMAC-SHA-256 API key hashing, per-agent rate limiting, input size limits on all fields, visibility-capped writes. Everything localhost by default.
**Agent management** — Add, list, update, and remove agents from the CLI. `agorai agent add` generates pass-keys, validates env vars, and configures both auth and adapter entries in one command.
**Internal agents** — Run agents inside the bridge process with `--with-agent`. Store-direct access, no HTTP round-trip. Perfect for always-on local models like Ollama.
**Session recovery** — Agents auto-reconnect with exponential backoff when the bridge restarts. No manual intervention needed.
## Docker
```bash
docker run -v ./agorai.config.json:/app/agorai.config.json -p 3100:3100 agorai/bridge
```
## Roadmap
| Version | Focus |
|---------|-------|
| **v0.2** | **Bridge — shared workspace, visibility, auth, 16 MCP tools** |
| v0.2.x | Security hardening, Docker, npm publish, session recovery, internal agents |
| **v0.3** | **SSE push notifications — real-time message delivery, 3-layer EventBus→Dispatcher→Client** |
| **v0.4** | **Metadata overhaul — bridgeMetadata/agentMetadata, confidentiality modes, high-water marks** |
| v0.4.x | Strict mode enforcement, discovery rules, access control |
| v0.5 | Discover, Decide, Deliver — capability catalog, task claiming, structured conversations, directed messages |
| v0.6 | Full-text search, archive conversations, Sentinel AI (auto-classification, redaction) |
| v0.7 | Web dashboard, human participants, A2A protocol support |
| v0.8+ | Enterprise — OAuth/JWT, RBAC, audit trail |
## Positioning
Agorai is **not** another agent framework. It's infrastructure — the collaboration layer that sits between your agents, regardless of which framework or model you use.
| | Agorai | CrewAI | AutoGen | LangGraph |
|---|---|---|---|---|
| Paradigm | Protocol-native collaboration | Role-based crews | Conversational | Pipeline/DAG |
| Protocol | MCP (open standard) | Custom | Custom | Custom |
| Models | Any (BYOM) | OpenAI-focused | OpenAI-focused | LangChain |
| Visibility | 4-level, store-enforced | None | None | None |
| Debate/consensus | Built-in | None | Basic | None |
| Local-first | Yes | Cloud-centric | Cloud-centric | Cloud-centric |
## License
AGPLv3. Dual licensing available for commercial use — reach out.
## Contributing
See [CONTRIBUTING.md](CONTRIBUTING.md).