Agent Workflow System
1139030773-cmd/agent-workflow-system
7 skills de IA colaborativas para converter ideias em tarefas com recuperação de sessão
Ver no GitHubSistema bilíngue de fluxo de trabalho com IA que transforma ideias vagas em tarefas executáveis através de 7 habilidades colaborativas (orientação, planejamento, execução, auditoria e encerramento). Inclui recuperação de sessão, auto-atualização via Git e constituição comportamental unificada para garantir consistência entre skills.
Instalação
npx convez-ai 1139030773-cmd/agent-workflow-system1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.
Métricas
Casos de uso
- Guiar usuários iniciantes de ideias vagas para tarefas claras e acionáveis
- Planejar e decompor projetos com critérios de aceitação e gates de decisão
- Executar correções de bugs, falhas de teste e debugging com TDD
- Auditar desvios de escopo e detectar task creep durante projetos
- Recuperar contexto de tarefas inacabadas entre sessões de trabalho
Veredito

Sistema robusto e bem estruturado para orquestração de workflows com IA, com foco em recuperação de contexto multi-sessão e separação clara de responsabilidades entre skills. Exige setup inicial em PowerShell/Bash e compreensão da arquitetura de hooks, mais adequado para equipes que precisam de continuidade em projetos longos.
README
A tradução para o português deste README ainda está na fila.
Agent Workflow System
A Chinese-English bilingual AI workflow system for Claude Code and Codex CLI. Turns vague ideas into executable tasks through 7 collaborative skills + behavioral constitution + session recovery. Beginner-friendly.
🚪 New Here? Start Here.
You don't need to understand how this system works. Just say what you want to do. The AI will guide you step by step — no jargon, no technical background required.
✨ Core Features
🔄 Session Recovery (v1.5.0+)
Close the window, the task survives. New window, same AI.
- CLAUDE.md — Auto-detects unfinished tasks on session start
- RESUME.md + context_snapshot — Checkpoint with decisions, preferences, and footgun prevention
- Open-ended recovery — No fixed options; AI presents status, you decide direction
- ⏱️ Time-interval awareness — Adjusts behavior based on how long since last session
- SessionEnd Hook — Auto-saves checkpoint + commits on window close
- SessionStart Hook — Auto
git pullon session start (24h throttle) - Cross-platform — PowerShell + Bash
🔧 Setup (2 minutes)
Step 1: Enable hooks — Add this to your .claude/settings.local.json:
{
"hooks": {
"SessionStart": [{
"matcher": "",
"hooks": [
{"type": "command", "command": "powershell.exe -File .claude/hooks/session-start.ps1", "timeout": 10}
]
}],
"SessionEnd": [{
"matcher": "",
"hooks": [
{"type": "command", "command": "powershell.exe -File .claude/hooks/session-end.ps1", "timeout": 15}
]
}]
}
}
Hook scripts are in hooks/ (copy to .claude/hooks/). Both .ps1 and .sh versions available.
Step 2: Run health check — scripts/health-check.ps1 check for system integrity scan.
🔁 Auto-Update Pipeline
You change system → Close window → Hook auto commit + push
↓
GitHub updated
↓
User opens window → Hook auto pull
🧠 7 Collaborative Skills
| Skill | Role | When |
|---|---|---|
workflow-system | Orchestrator | Entry routing to the right sub-skill |
newbie-guide | Guide | Vague goal → actionable task (Socratic questioning + optional background research) |
project-master | Planner | Task breakdown, decision gates, acceptance criteria |
debug-fixer | Executor | Errors / test failures / bugs (TDD minimal fix) |
learning-coach | Executor | Learning programming, design, languages, etc. |
drift-auditor | Auditor | Scope creep / context drift / task fork detection |
phase-closeout | Closer | Phase wrap-up, freeze state, generate resume token |
📜 Behavioral Constitution
15-chapter unified behavioral standard governing all skills:
- Role isolation — Planners don't code, executors don't make architectural decisions
- Five-level error correction — From self-check to mandatory human intervention
- Legal awareness — Three-layer risk alerts without giving legal advice
- Artifact handoff — 50k context → 5k artifact for lightweight inter-skill transfer
- Session recovery — Cross-session task continuity
- Task forking & convergence — Sub-tasks anchored to parent task context, auto-return on completion
- Four-question self-check — Before every change: logical? conflicts? important? right timing?
- Time-scale awareness — Tasks track created/last-active timestamps; stale decisions flagged for re-confirmation
CLAUDE.md Rules (v1.6.2 ~ v1.12.1):
- Verify by doing — Run commands, don't guess from docs
- Tone standards — Patient, equal, explain everything; no dismissive one-liners
- Task closeout rule — Never close a task without user confirmation
- Topic layer awareness — Recap what was done and where you stand in natural language
- Four-question self-check — Pre-implementation gate: logic / conflict / importance / timing
- Task registration rule — Every new task must be registered in task_stack before starting
- Task fork & converge — Sub-tasks inherit parent context; auto-return on completion
- Return-to-task format — After drift, resume with: goal / progress / next step
- Time-scale awareness — Real-world timestamps on tasks & decisions; auto-flag stale items after idle periods
🚀 Installation
Claude Code
Option 1: Ask Claude (Recommended)
In chat:
Install agent-workflow-system for me
Option 2: Manual
git clone https://github.com/1139030773-cmd/agent-workflow-system.git
cp -r agent-workflow-system/.claude ./ && cp agent-workflow-system/CLAUDE.md . && cp agent-workflow-system/RESUME.md .
Codex
Option 1: CLI One-liner (Recommended)
codex plugin install 1139030773-cmd/agent-workflow-system
Option 2: Desktop App — Install from URL
Settings → Plugins → Menu next to search → Install from URL:
https://raw.githubusercontent.com/1139030773-cmd/agent-workflow-system/main/plugins/agent-workflow-system/.codex-plugin/plugin.json
Option 3: Marketplace (requires PR merge)
codex plugin marketplace add https://github.com/hashgraph-online/awesome-codex-plugins.git
Then type /plugins in chat and find Agent Workflow System.
💡 Marketplace install depends on PR #202. Use Option 1 or 2 in the meantime.
Starter Prompts
After installation, try one of these:
| Scenario | Say |
|---|---|
| 🆕 Vague idea | "Start the newbie guide, help me turn my idea into tasks" |
| 📋 Manage a project | "Enter project master mode, help me manage this project" |
| 🐛 Something's broken | "Help me fix this bug" |
| 📖 Learn something | "I want to learn Python, make me a study plan" |
| 🔍 Feeling off-track | "Check if my progress has drifted" |
| 📦 Wrap up | "Help me do a phase closeout and organize what we did" |
💡 No commands to memorize. Speak naturally. The system auto-detects your language and intent.
📖 Version History
| Version | Date | Highlights |
|---|---|---|
| 1.12.1 | 2026-06-12 | 🛡️ Adversarial verification — independent agent checks for "should've changed but didn't" before release |
| 1.12.0 | 2026-06-12 | 🔍 Research gate — optional background research between newbie-guide and project-master |
| 1.11.1 | 2026-06-12 | health-check.ps1 encoding fix — 3 bugs, 100% health |
| 1.11.0 | 2026-06-12 | PROJECT.md — one-screen overview, auto-updated per phase |
| 1.10.3 | 2026-06-11 | Removed write-level gate — classification interference |
| 1.10.2 | 2026-06-11 | Full distribution — scripts + hooks included in plugin package |
| 1.10.0 | 2026-06-11 | One-click project archive — INDEX.md + NL recall |
| 1.9.2 | 2026-06-11 | Output self-check Q4 + phase-transition auto audit |
| 1.9.1 | 2026-06-11 | Project dashboard — health_score + active_tasks in RESUME.md |
| 1.9.0 | 2026-06-11 | Beginner mode — newbie entry + jargon filter + beginner channel |
| 1.8.2 | 2026-06-11 | Output self-check + write-level gate — prevent dev content leaking into workflow |
| 1.8.1 | 2026-06-11 | System health check + auto-cleanup — integrity scan on every session start |
| 1.8.0 | 2026-06-11 | Multi-window session management — .resume/ per-window files prevent conflicts |
| 1.7.7 | 2026-06-10 | Time-scale awareness — task timestamps + decision expiry + idle detection |
| 1.7.6 | 2026-06-10 | English adaptation — bilingual README + language auto-detect + starter prompts |
| 1.7.5 | 2026-06-09 | Skill exit protocol + Return-to-task format rule |
| 1.7.4 | 2026-06-09 | Mandatory format for drift → task return |
| 1.7.3 | 2026-06-09 | Sub-task anchoring: parent context = practice material |
| 1.7.2 | 2026-06-09 | Task registration rule + background recap on return |
| 1.7.1 | 2026-06-07 | Session closeout reminder + Version consistency validator |
| 1.7.0 | 2026-06-07 | CLAUDE.md split (dev rules / user rules) |
| 1.6.3 | 2026-06-07 | Task closeout rule |
| 1.6.2 | 2026-06-05 | Verify by doing + Tone standards |
| 1.6.1 | 2026-06-05 | Time-interval awareness + 7-step change protocol + Auto-update pipeline |
| 1.6.0 | 2026-06-05 | context_snapshot + Single-thread principle |
| 1.5.0 | 2026-06-05 | Session recovery mechanism, cross-platform hooks |
See CHANGELOG.md for full details.
🔗 Links
- GitHub: 1139030773-cmd/agent-workflow-system
- Codex:
codex plugin install 1139030773-cmd/agent-workflow-system - Community Marketplace: awesome-codex-plugins
- 中文文档
📄 License
MIT