Aller au contenu
TakoAPI
Retour aux skills

agent-team-kit

par ryancampbell

Communication
1.9kTéléchargements rapportés par clawskills.sh
5Étoiles
0Vues
0 mention J'aimeConnecte-toi pour aimer ce skill
—Pas encore noté
Ta note

Résumé

*A framework for self-sustaining AI agent teams.*.

Description

Ce que fait ce skill

A process kit for organizing AI agent teams with defined roles and a self-managing work queue. Agents discover, triage, and execute tasks through a 5-phase loop without requiring constant human direction. A heartbeat mechanism triggers periodic checks to keep work flowing when no human is present.

Removes the human bottleneck from task assignment by giving agents a shared, self-updating queue with unambiguous role ownership at each phase.

Quand l'utiliser

  • Running a multi-agent content pipeline without manual task assignment
  • Automatically triaging a backlog of discovered opportunities
  • Keeping a work queue populated via scheduled heartbeat checks
  • Delegating specialized tasks to sub-agents based on role definitions
  • Tracking active agent work status across a distributed team

Exemple de workflow

ENTRÉE

User asks: Set up an agent team to manage content discovery and publishing

AGENT
  1. 1

    Copy process template files to workspace root (INTAKE.md, ROLES.md, OPPORTUNITIES.md, BACKLOG.md, STATUS.md)

  2. 2

    Append heartbeat config to HEARTBEAT.md

  3. 3

    Customize ROLES.md with domain-specific role names and missions

  4. 4

    Scout agent discovers opportunities and logs them to OPPORTUNITIES.md

  5. 5

    Rhythm agent triages logged items and moves ready ones into BACKLOG.md

SORTIE

Agents pick up tasks from the ready queue autonomously, with hourly heartbeat checks maintaining loop momentum and surfacing stale or blocked work

Installation

clawhub install ryancampbell/agent-team-kit

Exécute la commande CLI, ou colle l'instruction dans Claude Code, Cursor, Codex ou l'agent de code que tu utilises.

Liens