Skip to content

erranthex/clawtasker

BranchesTags

Repository files navigation

ClawTasker — CEO Mission Control for Human + AI Teams

Version: 1.5.0 For: Human CEOs and AI Agents working together

ClawTasker is an open-source, local-first mission-control console for managing teams of human and AI agents. Sprint board, virtual office, mission planner, REST API — runs with python3 server.py, no Node.js required.

Both humans and AI agents are first-class users — every feature works through the GUI (for humans) and the REST API (for AI agents).

Screenshots

ClawTasker Dashboard

Installation

Quick Start (Standalone)

Open ui/dist/index.html directly in a browser. No server needed — runs entirely client-side with demo data.

Full Setup (with Server)

# 1. Set your API token
export CLAWTASKER_API_TOKEN="your-secret-token"

# 2. Start the server
python3 server.py

# 3. Open in browser
open http://localhost:3000

Build from Source

# Rebuild ui/dist/index.html from modular sources
python3 scripts/build_ui.py

# Run 137-check verification suite
python3 scripts/verify_build.py

No Node.js required — Python-only build toolchain.

Who Is This For?

ClawTasker is designed for any team where humans and AI agents work together:

Use Case Example
Software development AI coding agents + human tech lead
Business operations AI research agents + human CEO
Content & marketing AI writers + human creative director
Coaching & consulting AI analysis agents + human coach
Project management AI task agents + human PM

The platform is project-type agnostic — it works for software, coaching, business, marketing, or any domain.

Features

For Human Users (GUI)

Tab What You Can Do
Dashboard CEO overview — attention queue, mission brief, staffing, risk radar, sprint card
Team View org chart, add/remove agents, assign managers, edit roles and skills
Council Create/delete executive decisions with priority, status, and participants
Calendar View schedule, create events (meetings, reminders, standups), assign to agents
Board Sprint Kanban board, create tasks with full fields (DoD, epic link, story points)
Pipeline Full task backlog — filter by project/status, create and delete tickets
Approvals Review tasks in validation — approve or return with one click
Missions Create/edit/delete mission briefs with agents, success criteria, dependencies
Conversations Operator rail with directive history
Office Live 2D virtual office — watch agents work, take breaks, attend meetings
Access Project-agent access matrix
Requirements Create/edit/delete product requirements (P0-P3) — works for any project type
Test Cases Create/edit/run test cases linked to requirements
Appearance Dark/light mode, theme presets

For AI Agents (API)

Every GUI action has an API equivalent:

Action API Endpoint Method
Register as agent /api/agents/register POST
Send heartbeat /api/agents/heartbeat POST
Update agent profile /api/agents/update POST
Delete agent /api/agents/delete POST
Create task /api/tasks/create POST
Update task status /api/tasks/update POST
Delete task /api/tasks/delete POST
Comment on task /api/tasks/comment POST
Plan mission /api/missions/plan POST
Delete mission /api/missions/delete POST
Publish run result /api/openclaw/publish POST
Sync agent roster /api/openclaw/roster_sync POST
Get full state /api/snapshot GET
Real-time events /api/events/stream GET (SSE)
Integration contract /api/openclaw/contract GET

See API.md for the full specification.

Task Ticket Fields

Every task ticket supports:

  • Title and description
  • Assignee (human or AI agent)
  • Priority (P0 critical → P3 low)
  • Story points (Fibonacci: 1, 2, 3, 5, 8, 13)
  • Epic / Project link — connect tasks to projects/missions
  • Definition of Done — list of acceptance criteria
  • Status — Backlog → Ready → In Progress → Validation → Done
  • Mission link — connect to mission briefs
  • Comments (planned for v1.6.0)

AI Agent Onboarding

Prompt Pack

See docs/AGENT_PROMPTS.md for the Mission Control prompt pack:

  • openclaw/prompts/mission-control/01-existing-sub-agents-prompt.md
  • openclaw/prompts/mission-control/02-new-sub-agents-prompt.md
  • openclaw/prompts/mission-control/03-mission-control-orchestrator-agent-prompt.md

Onboarding Flow

  1. Register — POST /api/agents/register with name, role, skills, manager
  2. Read — GET /api/snapshot to understand current projects, tasks, and team
  3. Plan — POST /api/missions/update when staffing or dependencies change
  4. Work — POST /api/tasks/create and POST /api/tasks/update as work progresses
  5. Schedule — POST /api/calendar/events for meetings and reminders
  6. Communicate — surface blockers early, keep handoffs explicit

Integration Scripts

  • openclaw/register_agent.py — agent self-registration
  • openclaw/update_mission_plan.py — mission plan updates
  • openclaw/publish_status.py — task/heartbeat publisher
  • openclaw/publish_roster.py — roster sync

Architecture

Data Persistence

Current (v1.5.0): In-memory — data lives in JavaScript variables during the browser session. The server (server.py) maintains state in memory and serves it via /api/snapshot.

Planned (v1.6.0+): SQLite database for persistent storage. The data model is designed for easy migration:

  • agents table — id, name, role, skills, manager, status, last_heartbeat
  • tasks table — id, title, description, status, owner_id, project_id, priority, story_points, definition_of_done, mission_id, created_at
  • calendar_events table — id, day, time, title, agent, category, recurring
  • council_decisions table — id, title, summary, status, priority, date, participants
  • missions table — id, title, agents, success_criteria, dependencies, status
  • requirements table — id, title, description, priority, status, linked_tasks
  • test_cases table — id, title, linked_req, steps, expected, status, last_run

Workspace Structure

Each AI agent should create a workspace folder for their project work:

workspace/
├── <project-name>/
│   ├── docs/           — project documentation
│   ├── src/            — source code or deliverables
│   ├── tests/          — test files
│   ├── reports/        — status reports, analysis
│   └── README.md       — project overview

The ClawTasker platform stores its own data separately:

data/
├── agents.json         — agent registry
├── tasks.json          — task backlog
├── calendar.json       — calendar events
├── missions.json       — mission briefs
├── council.json        — council decisions
├── requirements.json   — product requirements
└── test-cases.json     — test cases

Module Architecture

ui/src/modules/         — 22 JS modules
├── data/constants.js   — all data constants (157KB)
├── state/store.js      — application state
├── lib/                — shared infrastructure
│   ├── router.js       — navigation
│   ├── dom.js          — DOM utilities
│   ├── theme.js        — dark/light mode
│   ├── counters.js     — dynamic counter updates
│   └── office-engine.js — canvas 2D game engine
├── views/              — 11 view modules
│   ├── dashboard.js, team.js, board.js, missions.js
│   ├── conversations.js, calendar.js, office.js, access.js
│   ├── appearance.js, requirements.js
│   └── council-pipeline-approvals.js
├── ui/                 — cross-cutting UI
│   ├── modals.js       — task creation modal
│   ├── onboarding.js   — platform onboarding
│   └── api.js          — SSE/API wiring
└── main.js             — entry point

Build: python3 scripts/build_ui.py → concatenates all modules into ui/dist/index.html

Project Structure

├── server.py                  — HTTP server + REST API
├── ui/
│   ├── dist/                  — Build output (auto-generated)
│   │   ├── index.html         — Self-contained app (~355KB)
│   │   └── assets/            — Portraits, sprites, textures, CSS
│   └── src/                   — Modular source (edit here)
│       ├── modules/           — 22 JS modules
│       ├── styles/            — CSS
│       ├── templates/         — HTML fragments
│       └── build-manifest.json
├── scripts/
│   ├── build_ui.py            — Build pipeline
│   └── verify_build.py        — CI verification (137 checks)
├── openclaw/                  — Agent integration scripts + prompts
├── schemas/                   — JSON schemas
├── docs/                      — Documentation + screenshots
├── API.md                     — Full REST API reference
├── CHANGELOG.md               — Release history
├── BILL_OF_MATERIALS.md       — Complete file manifest
└── REQUIREMENTS.md            — Product requirements

License

See LICENSE for details.

About

OpenClaw Mission Planner Task handler for human and AI agents

Topics

python rest-api project-management kanban ai-agents task-management mission-control virtual-office openclaw

Resources

Readme

License

MIT license

Code of conduct

Code of conduct

Contributing

Contributing

Security policy

Security policy
Activity

Stars

1 star

Watchers

0 watching

Forks

1 fork
Report repository

Releases 1

v1.5.0 Latest
Apr 1, 2026

Packages

 
 
 

Contributors

Languages