Skip to main content
Dispatcher

Agent-driven development, organized.

Give yourcoding agents aproper place to work.

A Dispatcher workflow in progress

Today’s dispatch3 agents working

In review

Make auth refresh impossible to race.

Codex finished · 12 tests passed · 2 files changed

Diff reviewedReady to merge

Running

Rebuild settings navigation.

Claude is editing 4 files

Up next

Your backlog, already in motion.

Pull tasks from Linear, GitHub, or Slack.

From request to reviewed pull request

One app. The whole loop.

01 / Plan

Turn requests into scoped work.

Bring in the issue, attach context, and define done before an agent writes a line.

02 / Run

Give every agent a clean room.

Isolated git worktrees keep parallel tasks moving without collisions or lost changes.

03 / Ship

Review the diff. Open the PR.

Approve a single pull request, or ship a full GitHub Stack of dependent PRs when the change is too big for one review.

The team request that becomes a merge

Slack to merged PR.

Mention @dispatcher in Slack and the rest stays in one window: a scoped task, an agent in an isolated worktree, review, then a pull request ready to merge.

  1. Mention

    @dispatcher in Slack

  2. Task

    Scoped work on the board

  3. Agent

    Runs in an isolated worktree

  4. Review

    Diff, terminal, and tests

  5. Merge

    Ship the PR from Dispatcher

Use the tools your team already trusts

Eleven harnesses. Zero lock-in.

  • Claude
  • Codex
  • Cursor
  • Kiro
  • OpenCode
  • Antigravity
  • Grok
  • Muse Code
  • Pi
  • omp
  • Hermes

Intake and shipping, already wired up

GitHub, Linear, Slack, and GitHub Stacks.

Also integrated

GitHub Stacks

Split large agent work into dependent stacked pull requests with GitHub’s gh stack flow. Dispatcher keeps the stack map, reviews, and merge controls in one place. No juggling branches in the terminal.

  • GitHub

    Import issues and open PRs

  • Linear

    Pull issues and projects onto the board

  • Slack

    Turn @dispatcher mentions into tasks

Context for agents, structure for teams

Task management and a knowledge base, built in.

01 / Tasks

Full task management.

Projects, boards, cycles, roadmaps, inbox, and review queues keep agent work organized from backlog through merge. The task is the prompt; the board is the queue.

02 / Knowledge

A knowledge base agents can use.

Specs, notes, and docs live next to the work and sync into the worktree, so agents start with your team’s context instead of an empty prompt.

Ready when your backlog is.

Put your agents to work without losing control.

Download Dispatcher
Dispatcher: Plan, build, and ship with any AI agent