Add OpenSwarm enterprise foundation

.env.enterprise.example

@@ -0,0 +1,20 @@
+OPENSWARM_ENTERPRISE=true
+ENVIRONMENT=development
+PORT=8080
+DATABASE_URL=postgresql+psycopg://openswarm:openswarm@postgres:5432/openswarm_enterprise
+REDIS_URL=redis://redis:6379/0
+ENTERPRISE_SECRET_KEY=replace-with-a-long-random-string
+# Generate with: python -c "from cryptography.fernet import Fernet; print(Fernet.generate_key().decode())"
+ENTERPRISE_ENCRYPTION_KEY=replace-with-a-fernet-key
+ACCESS_TOKEN_MINUTES=480
+CORS_ORIGINS=http://localhost:8080
+RATE_LIMIT_PER_MINUTE=120
+MAX_REQUEST_BYTES=10000000
+
+OPENAI_API_KEY=
+ANTHROPIC_API_KEY=
+GOOGLE_API_KEY=
+FAL_KEY=
+SEARCH_API_KEY=
+COMPOSIO_API_KEY=
+

.gitignore

@@ -80,6 +80,8 @@ cover/
 local_settings.py
 db.sqlite3
 db.sqlite3-journal
+*.db
+*.db-journal
 
 # Flask stuff:
 instance/
@@ -183,4 +185,4 @@ cython_debug/
 
 .agency_swarm/
 third_party/
-.claude/
+.claude/

README.md

@@ -6,6 +6,8 @@
 
 </div>
 
+> **Enterprise edition:** this fork adds a multi-tenant dashboard, REST API, RBAC, durable runs, encrypted provider keys, audit logs, usage records, and Docker deployment. See [docs/enterprise.md](docs/enterprise.md).
+
 **The fully open-source multi-agent system that does everything Claude Code can't.**
 
 Create polished slide decks, research reports, data visualizations, documents, images, and videos — all from a single prompt in your terminal. No platform, no UI, no setup hassles.
@@ -160,3 +162,4 @@ python server.py           # Runs on localhost:8080
 MIT — see [LICENSE](LICENSE).
 
 **Built with ❤️ by the team behind [Agency Swarm](https://github.com/VRSEN/agency-swarm)**
+<!-- Enterprise edition note: see docs/enterprise.md for the multi-tenant dashboard, REST API, RBAC, durable runs, encrypted secrets, audit logs, and Docker deployment added in this fork. -->

docker-compose.yml

@@ -8,3 +8,60 @@ services:
     volumes:
       - ./mnt:/app/mnt
       - ./uploads:/app/uploads
+
+  enterprise-api:
+    profiles:
+      - enterprise
+    build: .
+    command: python -u server.py
+    ports:
+      - "8080:8080"
+    env_file:
+      - .env.enterprise
+    environment:
+      OPENSWARM_ENTERPRISE: "true"
+      DATABASE_URL: postgresql+psycopg://openswarm:openswarm@postgres:5432/openswarm_enterprise
+      REDIS_URL: redis://redis:6379/0
+    depends_on:
+      - postgres
+      - redis
+    volumes:
+      - ./mnt:/app/mnt
+      - ./uploads:/app/uploads
+
+  enterprise-worker:
+    profiles:
+      - enterprise
+    build: .
+    command: python -u -m enterprise.worker
+    env_file:
+      - .env.enterprise
+    environment:
+      OPENSWARM_ENTERPRISE: "true"
+      DATABASE_URL: postgresql+psycopg://openswarm:openswarm@postgres:5432/openswarm_enterprise
+      REDIS_URL: redis://redis:6379/0
+    depends_on:
+      - postgres
+      - redis
+    volumes:
+      - ./mnt:/app/mnt
+      - ./uploads:/app/uploads
+
+  postgres:
+    profiles:
+      - enterprise
+    image: postgres:16
+    environment:
+      POSTGRES_DB: openswarm_enterprise
+      POSTGRES_USER: openswarm
+      POSTGRES_PASSWORD: openswarm
+    volumes:
+      - postgres-data:/var/lib/postgresql/data
+
+  redis:
+    profiles:
+      - enterprise
+    image: redis:7-alpine
+
+volumes:
+  postgres-data:

docs/enterprise.md

@@ -0,0 +1,44 @@
+# OpenSwarm Enterprise
+
+OpenSwarm Enterprise adds a multi-tenant control plane around the existing terminal-first OpenSwarm agents. The original `python swarm.py` CLI still starts the Agency Swarm TUI, and `python server.py` still starts the legacy Agency Swarm FastAPI integration unless `OPENSWARM_ENTERPRISE=true` is set.
+
+## Architecture
+
+- `enterprise/main.py` creates the enterprise FastAPI app, OpenAPI docs, dashboard routes, security headers, CORS, request-size limits, and rate limiting.
+- `enterprise/models.py` defines organizations, users, workspaces, agent configs, durable swarm runs, agent tasks, artifacts, encrypted provider keys, integration metadata, audit events, and usage records.
+- `enterprise/api.py` exposes the enterprise REST API.
+- `enterprise/orchestration.py` persists run lifecycle state and provides the adapter point for executing the live Agency Swarm workflow from a worker.
+- `enterprise/templates` and `enterprise/static` provide the web dashboard without adding a separate frontend build system.
+
+## Local Development
+
+```powershell
+pip install -r requirements.txt -r requirements-dev.txt
+Copy-Item .env.enterprise.example .env.enterprise
+$env:OPENSWARM_ENTERPRISE = "true"
+$env:ENTERPRISE_ENCRYPTION_KEY = python -c "from cryptography.fernet import Fernet; print(Fernet.generate_key().decode())"
+python scripts/migrate.py
+python scripts/seed_enterprise.py
+python server.py
+```
+
+Open `http://localhost:8080/login` and sign in with `admin@openswarm.local` / `ChangeMe123!`.
+
+## Docker
+
+```bash
+cp .env.enterprise.example .env.enterprise
+docker compose --profile enterprise up --build enterprise-api enterprise-worker postgres redis
+```
+
+## API
+
+OpenAPI is available at `/api/docs`.
+
+Important endpoints include `POST /api/auth/login`, `GET /api/me`, `GET /api/agents`, `POST /api/workspaces`, `POST /api/runs`, `GET /api/runs/{run_id}/stream`, `GET /api/artifacts/{artifact_id}`, `POST /api/integrations/secrets`, `GET /api/audit`, `GET /healthz`, and `GET /metrics`.
+
+## Security Notes
+
+Provider keys are encrypted with `ENTERPRISE_ENCRYPTION_KEY` and only masked values are returned by the API. Use a stable Fernet key in production, rotate it through a managed secret store, and never leave the example value in place.
+
+All API queries are scoped by `organization_id`. RBAC uses owner, admin, manager, member, and viewer roles.

enterprise/__init__.py

@@ -0,0 +1,2 @@
+"""Enterprise edition package for OpenSwarm."""
+

enterprise/agents.py

@@ -0,0 +1,15 @@
+ENTERPRISE_AGENTS = [
+    {
+        "key": "orchestrator",
+        "name": "Orchestrator",
+        "description": "Coordinates requests across the specialist OpenSwarm agents.",
+    },
+    {"key": "virtual_assistant", "name": "Virtual Assistant", "description": "Messaging, scheduling, task work, and Composio-powered tools."},
+    {"key": "deep_research", "name": "Deep Research", "description": "Evidence-based research with citations and balanced analysis."},
+    {"key": "data_analyst", "name": "Data Analyst", "description": "Structured data analysis, charts, and statistical modeling."},
+    {"key": "slides_agent", "name": "Slides Agent", "description": "HTML slide generation and PPTX export."},
+    {"key": "docs_agent", "name": "Docs Agent", "description": "Formatted Word and PDF document generation."},
+    {"key": "image_generation_agent", "name": "Image Generation Agent", "description": "Image generation and editing through configured providers."},
+    {"key": "video_generation_agent", "name": "Video Generation Agent", "description": "Video generation, editing, subtitles, and media composition."},
+]
+