Skip to content

deegalabs/clawhalla

Repository files navigation

ClawHalla

Enterprise AI agent orchestration platform built on OpenClaw.

Docker launcher + Mission Control dashboard + 9-agent hierarchy + smart contracts on Base L2.

What's Included

  • Docker setup — One-command OpenClaw installation with pre-configured workspace
  • Mission Control — 19-screen dashboard for agent orchestration and monitoring
  • Boards Engine — Multi-board Kanban with card detail, comments, history, drag-and-drop
  • Chat Engine — Multi-agent chat with party mode, voice input, streaming, session persistence
  • Content Pipelines — Multi-platform content publishing with Telegram approval flow
  • Telegram Integration — Inline approval buttons, media upload, webhook notifications
  • Memory/RAG System — Per-agent semantic search with local or cloud embeddings
  • Autopilot — Goal-driven autonomous execution with human feedback loop
  • 9 AI agents — Norse mythology-themed hierarchy across 4 squads
  • Agent Factory — Create and configure agents with per-agent memory modes
  • Squad Pack system — Install pre-configured agent teams with one click
  • Secret Vault — AES-256-GCM encrypted credential storage
  • Notification system — SSE with auto-reconnect, exponential backoff, sound alerts
  • Terminal — Sandboxed shell with command blocklist and cwd restriction
  • Smart Contracts — AgentRegistry, LicenseNFT, Marketplace on Base L2
  • Full-text search — SQLite FTS5 across 200+ workspace documents
  • Real-time updates — SSE + chokidar file watching for live dashboard
  • AI-AGIL methodology — Structured multi-agent development framework

Quick Start

git clone https://github.com/deegalabs/clawhalla.git
cd clawhalla
cp .env.example .env
# Edit .env with your Anthropic API key
docker compose up -d --build
docker compose exec clawhalla bash
openclaw onboard

Mission Control

After OpenClaw is running, start Mission Control:

# Inside the container:
cd ~/repos/clawhalla/apps/mission-control
pnpm install
pnpm dev --hostname 0.0.0.0 --port 3000

Access at http://localhost:3333

Screens

Screen Description
Dashboard System health, agent stats, real-time activity feed (SSE)
Boards Multi-board Kanban engine with card detail, comments, history
Calendar Cron jobs and scheduled tasks
Projects Project cards + Git push/pull panel
Pipeline Content publishing — multi-platform drafts with approval flow
Chat Multi-agent chat with party mode, voice input, streaming
Approvals CEO approval gates and decision history
Memory Agent memory with RAG semantic search, daily journal, FTS5
Docs Searchable document browser with category filters
Content Post editor with LinkedIn integration + checklist
Council R&D council sessions — automated research memos
Office Live agent status and session monitoring
Terminal Sandboxed terminal with command blocklist
Autopilot Goal-driven autonomous execution with feedback loop
Marketplace Squad packs + wallet connect (Base L2)
Settings Secret vault, memory/RAG config, per-agent memory modes
Onboarding Setup wizard for first-time users
Pipeline Build status and CI/CD monitoring
Feedback Continuous learning and agent feedback system

API Routes (40+)

Core
  GET  /api/health              — Gateway health check (debounced)
  GET  /api/sse                 — Server-Sent Events stream
  GET  /api/search?q=query      — FTS5 full-text search

Boards Engine
  CRUD /api/boards              — Board management
  CRUD /api/boards/:id/cards    — Card CRUD with SQL-level filtering
  GET  /api/boards/:id/cards/:id/comments — Card comments
  GET  /api/boards/:id/cards/:id/history  — Card history

Tasks & Projects
  CRUD /api/tasks               — Legacy task management
  CRUD /api/projects            — Project management
  CRUD /api/epics               — Epic tracking
  GET  /api/sprints             — Sprint management

Chat & AI
  POST /api/chat                — Streaming chat (rate-limited)
  CRUD /api/chat/sessions       — Chat session persistence
  POST /api/dispatch            — Agent task dispatch (rate-limited)
  POST /api/council/session     — R&D council sessions

Content & Knowledge
  GET  /api/activities          — Activity log (paginated)
  GET  /api/memory              — Memory entries
  GET  /api/docs                — Document scanner (capped)
  CRUD /api/content/drafts      — Content drafts
  CRUD /api/content/pipelines   — Content pipelines
  POST /api/content/publish      — Multi-platform content publishing
  GET  /api/content/platforms    — Available platform adapters
  POST /api/telegram/webhook     — Telegram bot webhook (approvals, media)
  GET  /api/memory/config        — Memory/RAG provider configuration
  GET  /api/memory/rag?q=query   — Semantic search across agent memory
  POST /api/memory/rag           — Trigger memory reindex
  GET  /api/notifications       — Notification system

Agents & System
  GET  /api/agents/health       — Agent health status
  CRUD /api/agents/factory      — Agent Factory
  GET  /api/org-structure       — Agent hierarchy (YAML)
  CRUD /api/packs               — Squad Pack installer
  GET  /api/gateway/sessions    — Live agent sessions
  GET  /api/gateway/crons       — Cron job list
  CRUD /api/crons               — Cron management
  POST /api/squads/create        — Squad creation with templates

Security & Config
  CRUD /api/vault               — Secret vault (AES-256-GCM)
  POST /api/vault/reveal        — Decrypt secret
  POST /api/terminal            — Sandboxed terminal
  CRUD /api/settings            — System settings
  POST /api/reset               — System reset (auth required)
  GET  /api/auth/session        — Session token

External
  GET  /api/git                 — Git repo status
  POST /api/git                 — Git push/pull
  CRUD /api/linkedin            — LinkedIn integration
  POST /api/feedback            — Continuous learning
  CRUD /api/approvals           — Approval workflow

Agent Hierarchy

Daniel (CEO, human)
  └── Claw 🦞 — Chief Orchestrator (Opus 4.6)
        ├── Dev Squad
        │     ├── Thor ⚡ — Tech Lead (Sonnet 4.6)
        │     ├── Freya ✨ — Senior Developer (Sonnet 4.6)
        │     └── Tyr ⚖️ — Security Auditor (Sonnet 4.6)
        ├── Personal Squad
        │     ├── Frigg 👑 — Coordinator/PA (Sonnet 4.6)
        │     └── Mimir 🧠 — Knowledge Curator (Sonnet 4.6)
        ├── Social Squad
        │     ├── Saga 🔮 — Content Strategist (Sonnet 4.6)
        │     └── Bragi 🎭 — Content Creator (Sonnet 4.6)
        └── Vidar ⚔️ — Blockchain Architect (Sonnet 4.6)

Smart Contracts (Base L2)

Contract Purpose
AgentRegistry On-chain template registry with creator attribution
LicenseNFT ERC-721 license tokens (transferable, revocable)
Marketplace Purchase flow with royalty splits (2.5% platform fee)

Audited by Tyr (AI security auditor). All findings resolved.

Tech Stack

  • Runtime: Docker + OpenClaw Gateway
  • Dashboard: Next.js 16, TypeScript strict, Tailwind CSS v4
  • Database: SQLite + Drizzle ORM (23 tables, 11 indexes) + FTS5
  • Real-time: SSE + chokidar file watching
  • Security: CSP, CORS middleware, rate limiting, XSS prevention, AES-256-GCM vault
  • Wallet: wagmi + viem (Base L2, Base Sepolia, Ethereum mainnet)
  • Contracts: Solidity 0.8.24, Foundry, Base L2
  • Search: SQLite FTS5 with porter stemming

Project Structure

clawhalla/
├── apps/mission-control/   — Next.js dashboard (19 screens, 40+ API routes)
├── agents/                 — Agent persona files (SOUL.md, AGENTS.md, skills/)
├── contracts/              — Solidity smart contracts (Foundry)
├── docker/                 — Docker configuration
├── docs/                   — Architecture, roadmap, security docs
├── scripts/                — Setup and utility scripts
├── squads/                 — Squad templates (dev, blockchain, clop-cabinet)
├── workspace-template/     — Enterprise workspace structure + AI-AGIL methodology
├── Dockerfile
├── docker-compose.yml
└── README.md

Business Model

Tier What Price
Free (open source) Docker + MC (19 screens) + 3 agents $0
Pro packs Squad packs, unlimited agents/projects $49/mo
Premium skills Blockchain auditor, DeFi monitor, etc. $20-100/skill
Marketplace Agent templates as NFT licenses (Base L2) Creator-set pricing

License

MIT

Links

About

Enterprise AI agent orchestration platform built on OpenClaw.

Resources

License

Contributing

Security policy

Stars

Watchers

Forks

Packages

 
 
 

Contributors