AgentFlow Skill · Devops

AgentFlow: Autonomous AI Development Pipeline

Orchestrate multi-worker Claude Code pipelines through your Kanban board.

7-stage pipeline with deterministic quality gates, crash-proof recovery, and per-task cost tracking. Download now.

  • Kanban
  • Claude Code
  • Orchestration
  • Quality Gates
  • Autonomous Pipeline
  • Cost Tracking

About This Skill

AgentFlow transforms your Kanban board into a fully autonomous AI development pipeline with 7 stages (Backlog, Research, Build, Review, Test, Integrate, Done), deterministic quality gates, and crash-proof orchestration. Manage multiple Claude Code workers with full observability from your phone.

Quick Start

  1. 1Connect your Kanban board (Asana, GitHub Projects, or Linear)
  2. 2Define your pipeline stages and quality gates
  3. 3Dispatch tasks and let AgentFlow orchestrate AI workers automatically
Example Command
agentflow run --board asana --project 'Sprint 42' --stages backlog,research,build,review,test,integrate,done

Core Capabilities

7-Stage Kanban Pipeline

Tasks flow through Backlog, Research, Build, Review, Test, Integrate, and Done with deterministic gates at each stage.

Stateless Orchestrator

Crontab-driven one-shot sweep runs every 15 minutes with no daemon or session dependency. State lives in your PM tool for crash-proof recovery.

Deterministic Before Probabilistic

Hard gates (tsc, eslint, tests) run before AI review, ensuring code quality before probabilistic AI analysis.

Per-Task Cost Tracking

Track costs for each AI worker and task, with full pipeline observability from your phone or Kanban board.

Usage Examples

Before

Manual handoffs between developers, AI tools, and QA — fragmented, error-prone, and slow.

After

Fully automated pipeline with human override at any point by dragging a card. Complete observability from your phone.

Input

Fix a bug reported in GitHub Issues

Output

AgentFlow picks up the issue, dispatches a Claude Code worker to research and fix, runs quality gates, and creates a PR — all tracked on your Kanban board.

Input

Run a solo project with autonomous task dispatch

Output

AgentFlow manages multiple AI workers across build, review, and test stages, with cost tracking and crash-proof recovery if your session restarts.

SKILL.md

---
name: agentflow
description: "Orchestrate autonomous AI development pipelines through your Kanban board (Asana, GitHub Projects, Linear). Manages multi-worker Claude Code dispatch, deterministic quality gates, adversarial review, per-task cost tracking, and crash-proof pipeline execution."
risk: safe
source: community
date_added: "2026-04-02"
---

# AgentFlow

## Overview

AgentFlow turns your existing Kanban board into a fully autonomous AI development pipeline. Instead of building custom orchestration infrastructure, it treats your project management tool (Asana, GitHub Projects, Linear) as a distributed state machine — tasks move through stages, AI agents read and write state via comments, and humans intervene through the same UI they already use.

The result is complete pipeline observability from your phone, free crash recovery (state lives in your PM tool, not in memory), and human override at any point by dragging a card.

## When to Use This Skill

- Use when you need to orchestrate multiple Claude Code workers across a full development lifecycle (build, review, test, integrate)
- Use when you want deterministic quality gates (tsc/eslint/tests) before AI review on AI-generated code
- Use when you want full pipeline visibility from your Kanban board or phone
- Use when running a solo or team project that needs autonomous task dispatch with cost tracking
- Use when you need crash-proof orchestration that survives session restarts

## Core Concepts

### 7-Stage Kanban Pipeline

Tasks flow through: Backlog, Research, Build, Review, Test, Integrate, Done. Each stage has specific gates. The Kanban board IS the orchestration layer — no separate database, no message queue, no custom infrastructure.

### Stateless Orchestrator

A crontab-driven one-shot sweep runs every 15 minutes. No daemon, no session dependency. If it crashes, the next sweep picks up where it left off because all state lives in your PM tool.

### Deterministic Before Probabilistic

Hard gates 

Frequently Asked Questions

FAQ

Which project management tools does AgentFlow support?
AgentFlow works with Asana, GitHub Projects, and Linear. It treats your Kanban board as a distributed state machine, so tasks move through stages and AI agents read/write state via comments.
Who is AgentFlow designed for?
AgentFlow is for solo developers and teams who want to orchestrate multiple Claude Code workers across a full development lifecycle — build, review, test, integrate — with full pipeline visibility from their Kanban board or phone.
How is AgentFlow different from other AI orchestration tools?
Unlike custom orchestration infrastructure, AgentFlow uses your existing Kanban board as the orchestration layer — no separate database, no message queue, no custom infrastructure. It's crash-proof because state lives in your PM tool, not in memory.
What programming languages and tools does AgentFlow support?
AgentFlow supports any language that Claude Code can handle. Deterministic quality gates include tsc (TypeScript), eslint, and test frameworks. The pipeline is language-agnostic at the orchestration level.
What results can I expect from using AgentFlow?
You get a fully autonomous 7-stage pipeline with deterministic quality gates, per-task cost tracking, and crash-proof recovery. Tasks flow from Backlog to Done without manual intervention, and you can override any step by dragging a card.

Discussion

Discussion

0 comments
U

Trigger Phrases

Use these phrases to activate this skill in your AI coding assistant:

autonomous AI pipelineKanban orchestrationClaude Code multi-workercrash-proof pipelinedeterministic quality gatesper-task cost tracking
AgentFlow: Autonomous AI Development Pipeline for Developers