Open Station

Task management system for coding AI agents. Pure convention — markdown specs + skills, zero runtime dependencies.

What It Is

Open Station gives coding AI agents a structured way to receive, execute, and complete tasks. Everything is plain markdown with YAML frontmatter — no runtime, no database, no dependencies. Drop it into any project and agents can discover tasks, follow a defined work process, and store artifacts, all through file conventions.

Install

curl -fsSL https://raw.githubusercontent.com/leonprou/openstation/main/install.sh | bash

This creates the vault directories, installs skills and the manual, sets up symlinks for slash-command discovery, and adds an Open Station section to your CLAUDE.md.

Options:

Flag	Description
--local PATH	Copy from a local clone instead of downloading
--no-agents	Skip installing example agent specs

Quick Start

1. Create a task

/openstation.create Add input validation to the signup form

2. Set it ready

/openstation.ready 0001-add-input-validation agent:researcher

3. Dispatch an agent

claude --agent researcher

The agent finds its ready tasks, follows the manual, executes the work, and sets status: review when done.

4. Mark done

/openstation.done 0001-add-input-validation

Moves the task symlink to tasks/done/.

Vault Structure

docs/              — Project documentation (lifecycle, task spec, README)
tasks/             — Lifecycle buckets (contain symlinks, not real folders)
  backlog/         —   Not yet ready for agents
  current/         —   Active work (ready → in-progress → review)
  done/            —   Completed tasks
artifacts/         — Canonical artifact storage (source of truth)
  tasks/           —   Task folders (canonical location, never move)
  research/        —   Research outputs
  specs/           —   Specifications & designs
agents/            — Agent specs (identity + skill references)
skills/            — Agent skills (not user-invocable)
commands/          — User-invocable slash commands

When installed into another project via install.sh, these are placed under .openstation/.

Architecture

                        ┌──────────┐
                        │ CLAUDE.md│
                        └────┬─────┘
                             │
                  references │
             ┌───────────────┤
             ▼               ▼
      ┌────────────┐  ┌───────────┐
      │lifecycle.md │◄─┤task.spec.md│
      └──┬──────┬──┘  └───────────┘
         │      │
         │      └───────────┐
         ▼                  ▼
┌─────────────────────────────┐
│  skills/                    │
│  openstation-execute/       │──► lifecycle.md
└────────────────┬────────────┘    task.spec.md
                 │                 /openstation.create
      skills:    │                 /openstation.done
      execute    │
         ┌───────┴───────┐
         ▼               ▼
  ┌──────────┐    ┌──────────┐
  │researcher│    │  author  │
  └──────────┘    └──────────┘

┌─────────────────────────────────────────┐
│  commands/                               │
│  openstation.create.md  ──► lifecycle.md │
│  openstation.done.md    ──► lifecycle.md │
│  openstation.update.md  ──► lifecycle.md │
│  openstation.list.md                     │
│  openstation.dispatch.md──► agents/      │
└─────────────────────────────────────────┘

Doc	Purpose
task.spec.md	The shape — schema, naming, format
lifecycle.md	The state machine — transitions, ownership, artifacts
execute skill	The agent playbook — discovery, execution, completion

Commands

Command	Description
/openstation.create	Create a new task spec from a description
/openstation.list	List all tasks with status, agent, and dates
/openstation.show	Show full details of a single task
/openstation.ready	Promote a task from backlog to ready
/openstation.update	Update task frontmatter fields
/openstation.done	Mark a task done and move it to done/
/openstation.reject	Reject a task in review and mark it failed
/openstation.dispatch	Preview agent details and show launch instructions

License

MIT