Skip to content
Skills Library
AI & agents · Automation

dispatch

/dispatch

Use when the user wants work to begin or to move on after a merge. Starts the task issues that are ready - no open dependency, a free slot for their role - with the project's starter (agent:<role> label on GitHub, pipeline on GitLab, a configured command such as an orchestrator, on any tracker), but only after showing the plan and getting the user's yes. Safe at any time - it never starts what is running, blocked, on hold or on-demand.

Repository
agent-team ↗
Licence
MIT
Last repository change
28.10.2026

Assessment and security

100/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
1
Scripts
0
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 28.10.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

Use when the user wants work to begin or to move on after a merge. Starts the task issues that are ready - no open dependency, a free slot for their role - with the project's starter (agent:<role> label on GitHub, pipeline on GitLab, a configured command such as an orchestrator, on any tracker), but only after showing the plan and getting the user's yes. Safe at any time - it never starts what is running, blocked, on hold or on-demand.

Arguments

/dispatch [--per-role N] [--max N] [--milestone M1 | --all]
Original author description

Use when the user wants work to begin or to move on after a merge. Starts the task issues that are ready - no open dependency, a free slot for their role - with the project's starter (agent:<role> label on GitHub, pipeline on GitLab, a configured command such as an orchestrator, on any tracker), but only after showing the plan and getting the user's yes. Safe at any time - it never starts what is running, blocked, on hold or on-demand.

How it works

  1. 1. Preconditions
  2. 2. Show the plan
  3. 3. Start
  4. In Codex or OpenCode
Original SKILL.md headings · 46 lines, 298 words
SKILL.md preview
---
name: dispatch
description: Use when the user wants work to begin or to move on after a merge. Starts the task issues that are ready - no open dependency, a free slot for their role - with the project's starter (agent:<role> label on GitHub, pipeline on GitLab, a configured command such as an orchestrator, on any tracker), but only after showing the plan and getting the user's yes. Safe at any time - it never starts what is running, blocked, on hold or on-demand.
argument-hint: "[--per-role N] [--max N] [--milestone M1 | --all]"
---

Start the task issues that can run now. Talk to the user in their language.

## 1. Preconditions

- The script: `scripts/dispatch.py` in the project if it exists together
  with `scripts/agentteam/`, otherwise
  `${CLAUDE_PLUGIN_ROOT}/template/scripts/dispatch.py`. Run it from the
  project root.
- A reachable tracker: `python3 <scripts>/tracker.py check`. It also prints
  the *starter*, which decides what "start" means: the `agent:<role>` label
  (GitHub), the agent pipeline (GitLab), the project's `start_command` (for
  example an orchestrator) or a printed instruction. Tell the user which one
  applies before anything starts.
- On GitHub the `agent:<role>` labels must exist (`gh label list`); if they
  do not, offer to run `scripts/create-labels.sh` first.

## 2. Show the plan

Run `python3 <script> $ARGUMENTS` (without `--apply`) and explain the output:

- the scope (by default the first open milestone that still has tasks);
- what is running now, and what would start;
- what waits, and for what: an open dependency, or a role already busy;
- what is never started automatically: optimizer and security (on-demand),
  issues with the `hold` label, "Design:" issues.

Each start is an agent run, so it costs tokens or subscription quota. The
defaults are one task per role and three in total; the user can change them
with `--per-role` and `--max`.

## 3. Start

Only after the user's yes, run the same command with `--apply`. Report what
started and what will start next, as the running tasks are merged.

To make this happen by itself on GitHub every time an issue closes, point the
user to `.github/workflows/dispatch.yml`: it needs the `AGENT_AUTODISPATCH`
variable and the `AGENT_DISPATCH_TOKEN` secret described at its top.

## In Codex or OpenCode

`${CLAUDE_PLUGIN_ROOT}` is the plugin folder in Claude Code. In Codex or
OpenCode, which load these skills from `~/.agents/skills/`, use the folder where
the agent-team repository is cloned (`$AGENT_TEAM_HOME`) instead.

Versions

RepositoryScoreLicence
This page · AlbertoRiorda95/agent-team100MIT
jeremylongshore/tons-of-skills-marketplace100MIT
Sagargupta16/foreman100MIT
LandonSchropp/agent-toolkit100MIT
bendrucker/claude100MIT
renzrollon/interlock100MIT
epilande/ccmux100MIT
bengous/claude-code-plugins100MIT
tylerjrbuell/Wardens100MIT

Database snapshot: 13.11.2026 · package version 557d7a16fb7e (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 90create-issues
    Use when the roadmap and ready specs in docs/design/ should become issues. Creates the milestones and issues (GitHub, GitLab, local files, Jira, Linear or Redmine) from the roadmap and the ready specs, but only after showing the plan and getting the user's yes; nothing is…
  • 100dispatch-agents
    Use when a multi-task plan or backlog has dependency-ordered tasks that can run in parallel without file conflicts.
  • 100dispatch-first
    Use when a session has the dispatch command, before you plan, file an issue, ask a human, post a finding, or start work. Someone may track that work already or hold a decision on it.

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.