https://github.com/pcliangx/appgenesisforge
Production-ready AI Agent Team template for Claude Code — 14 specialists, Agile+Scrum+TDD enforced via skill/hook/DoD 3-layer (not wiki), PRD→UAT 6 stage gates, 4-layer hook defense, multi-LLM (DeepSeek/Doubao/Qwen/MiniMax), WeChat Mini Program, trilingual README (中文/EN/日本語). Ship a feature overnight.
https://github.com/pcliangx/appgenesisforge
agent-team agile-scrum ai-agents claude claude-code deepseek doubao fastapi llm minimax multi-agent prompt-engineering python qwen rag react sub-agents tdd template wechat-miniprogram
Last synced: 22 days ago
JSON representation
Production-ready AI Agent Team template for Claude Code — 14 specialists, Agile+Scrum+TDD enforced via skill/hook/DoD 3-layer (not wiki), PRD→UAT 6 stage gates, 4-layer hook defense, multi-LLM (DeepSeek/Doubao/Qwen/MiniMax), WeChat Mini Program, trilingual README (中文/EN/日本語). Ship a feature overnight.
- Host: GitHub
- URL: https://github.com/pcliangx/appgenesisforge
- Owner: pcliangx
- License: mit
- Created: 2026-04-28T03:36:39.000Z (about 2 months ago)
- Default Branch: main
- Last Pushed: 2026-05-25T07:16:49.000Z (28 days ago)
- Last Synced: 2026-05-25T07:23:17.630Z (28 days ago)
- Topics: agent-team, agile-scrum, ai-agents, claude, claude-code, deepseek, doubao, fastapi, llm, minimax, multi-agent, prompt-engineering, python, qwen, rag, react, sub-agents, tdd, template, wechat-miniprogram
- Language: Python
- Homepage:
- Size: 62.8 MB
- Stars: 8
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.en.md
- Changelog: CHANGELOG.md
- License: LICENSE
Awesome Lists containing this project
README
# App Genesis Forge · Your AI Coworker Team
> **Code the Origin, Forge the App.**
> 14 AI coworkers · Agile + Scrum + TDD backbone · ship a feature overnight · watch it all in one browser tab.
[](https://github.com/pcliangx/AppGenesisForge/generate)
[](https://github.com/pcliangx/AppGenesisForge/releases)
[](https://opensource.org/licenses/MIT)
[](https://claude.com/claude-code)
[](#-core-thinking-agile--scrum--tdd)
[](.claude/agents/)
[](.claude/commands/)
[](.claude/skills/)
[](.claude/skills/agf-wiring-multi-llm-sdk/)
[](.claude/standards/miniapp.md)
[](.claude/standards/security.md)
[](https://github.com/pcliangx/AppGenesisForge/stargazers)
[](https://github.com/pcliangx/AppGenesisForge/pulls)
[简体中文](README.md) · **English**
> 📖 **Reading path (Setup → Detail → Wrap)**:
> - **Setup (conclusion first)** ── [💡 Conclusion](#-what-it-is-conclusion-first) → [🧠 Philosophy](#-core-thinking-agile--scrum--tdd)
> - **Detail (3 pillars + onboarding)** ── [🏢 Organization](#-organization--who) → [🔄 Process](#-process--how) → [⚙️ Mechanism](#-mechanism--enforcement) → [⏱ Day 1](#-day-1) → [🛠️ Tools](#-self-service)
> - **Wrap (closing)** ── [🔍 Comparison](#-vs-others) → [✨ Summary](#-summary)
---
## 💡 What It Is (Conclusion First)
**AppGenesisForge** is an **AI team scaffold** on [Claude Code Agent Teams](https://code.claude.com/docs/en/agent-teams) — not a prompt collection, not a roster of agent personalities, but a **virtual AI team with Agile + Scrum + TDD backbone, defined workflow + standards, and acceptance gates**.
### Core conclusion (one sentence)
> **Agile + Scrum + TDD principles are wired into the skill / hook / DoD enforcement path — never relying on agent self-discipline.**
### 6 facts
- **🧠 Methodology backbone is non-negotiable** — Agile + Scrum + TDD ([see §Philosophy](#-core-thinking-agile--scrum--tdd))
- **🏢 Clear organization** — 14 AI coworkers with role contracts ([see §Organization](#-organization--who))
- **🔄 Closed-loop process** — PRD → UAT 6 stage gates, no skipping on failure ([see §Process](#-process--how))
- **🚀 Controllable parallelism** — Multi-instance Worker Pool ([ADR-001](docs/adr/001-multi-instance-worker-pool.md)): ≥2 same-type pending tasks → Dev / Reviewer / QA three-layer auto fan-out; PL fans-in with `agf-matrix.sh`
- **⚙️ Mechanism backstop** — skill enforcement / hook block / DoD checklist, three layers ([see §Mechanism](#-mechanism--enforcement))
- **🇨🇳 China-native ecosystem** — DeepSeek / Doubao / Qwen / MiniMax + WeChat Mini Program ([see §Tools](#-self-service))
### Boundary
✅ **Built for**: React + FastAPI + Postgres · AI Agent / RAG / multimodal · WeChat Mini Programs (native + Taro fallback) · full PRD → UAT loop
❌ **Not built for**: native iOS / Android · desktop GUI · large-model training
> 📝 **Default backend is Python (FastAPI + SQLAlchemy + Alembic)**, frontend React + Vite, **extensible to Java / Go / Rust** and other backend stacks — OpenAI-compatible LLM wiring / Agile + Scrum + TDD principles / 4-layer hook defense / 14-coworker role contracts are all language-agnostic. To swap stacks, `tech-lead` rewrites [`ADR-000`](.claude/skills/agf-writing-adr/SKILL.md) + the Python examples in [`backend-dev`](.claude/agents/backend-dev.md) / [`testing.md`](.claude/standards/testing.md) / `agf-wiring-multi-llm-sdk`.
---
## 🧠 Core Thinking: Agile + Scrum + TDD
AI agents have two natural weaknesses: **lost context + over-confidence**. Scrum stage gates + Agile iterative feedback + TDD red-green-refactor are the mature disciplines against both, and this project **lands them across three layers — Organization + Process + Mechanism**. This section covers WHAT/WHY; the next three cover HOW.
Full decision log: [`product-workflow.md §4.4`](./docs/product-workflow.md#44-agile--scrum-心法采用清单--词汇替代) + [`team-capability-map.md §1.6`](./docs/team-capability-map.md#16-scrum-词汇--本项目载体feature-流变体).
### Principles checklist (all mandatory / three-layer landing)
| Agile / Scrum concept | Project carrier | Landing layer |
|---|---|---|
| **TDD red-green-refactor** | `superpowers:test-driven-development` + DoD order rule | ⚙️ Mechanism |
| **AC (Acceptance Criteria)** | PRD §4 table (P0/P1) + dispatch must quote AC verbatim | 🔄 Process + ⚙️ Mechanism |
| **DoD (Definition of Done)** | [`ac-lifecycle.md`](.claude/standards/ac-lifecycle.md) 9-item checklist | ⚙️ Mechanism |
| **DoR (Definition of Ready)** | `brainstorming` done + AC verifiable in ≤30s + Open Q has Owner | 🔄 Process |
| **User Story (INVEST)** | PRD §3 table + US↔AC ID cross-reference | 🔄 Process |
| **Product Owner + Scrum Master** | `product-lead` (PO+SM merged) | 🏢 Organization |
| **Dev Team (cross-functional)** | 5 dev + 1 Designer + 2 Reviewer + 2 QA + 2 Post-launch = 12 + PL/TL = 14 | 🏢 Organization |
| **Sprint** | One feature = PRD → UAT signoff loop | 🔄 Process |
| **Sprint Planning** | `superpowers:writing-plans` (async) | 🔄 Process |
| **Daily Standup** | `progress/.md` + SendMessage (async) | ⚙️ Mechanism |
| **Sprint Review** | UAT business signoff | 🔄 Process |
| **Sprint Retrospective** | `/agf-release-retro vX.Y.Z` (mandatory for MAJOR/MINOR) | 🔄 Process |
### Scrum vocabulary → project carrier (substitution map)
| Scrum term | Project substitute | Reason |
|---|---|---|
| **Sprint** (timed iteration) | Feature stream | Agent Team is continuously async; no sync cadence |
| **Product Backlog** | `docs/prd/` files with Status=Draft/Review | File-based, git-trackable |
| **Sprint Backlog** | `~/.claude/tasks//*.json` | PL maintains via `TaskCreate/Update`; viewed via `/agf-tasks` |
| **Story Points** | PRD `Estimated effort tier: S/M/L` | Tied to token budget tiers; more precise than abstract points |
| **Velocity / Burndown** | Not tracked | No timed iterations → velocity meaningless; ≥3 retros later, cycle-time trend is evaluated (implicit form) |
| **Daily Standup** | `progress/.md` + SendMessage | Async standup + hook backstop |
---
## 🏢 Organization / WHO
> Scrum 3-role mapping + 14 AI coworkers. **Self-managed, cross-functional** — devs self-run Unit + SIT, Reviewers don't touch source, QA doesn't sign business off.
### Scrum 3-role mapping
| Scrum role | Project carrier | Notes |
|---|---|---|
| **Product Owner** | [`product-lead`](.claude/agents/product-lead.md) | PRD / priority P0/P1 / UAT business signoff |
| **Scrum Master** | `product-lead` (PO+SM merged) | Workflow orchestration / skill enforcement / retro chairing. Agent has no human conflict of interest; merging is reasonable. |
| **Development Team** | 12 execution / review / QA / post-launch roles (see below) | Self-managed; devs self-run Unit + SIT; ≥2 same-type pending tasks auto fan-out into a Pool ([ADR-001](docs/adr/001-multi-instance-worker-pool.md)) |
| **Technical Advisor** (not a Scrum role) | [`tech-lead`](.claude/agents/tech-lead.md) | Only intervenes for missing baseline / new tech selection / architecture risk escalation |
### The 14 AI coworkers
| # | Coworker | Personality + responsibility |
|---|---|---|
| 🟠 001 | [`product-lead`](.claude/agents/product-lead.md) | **PO + SM** — never codes, only signs. PRD / task assignment / UAT signoff / retro chair |
| 🔵 002 | [`tech-lead`](.claude/agents/tech-lead.md) | Battle-scarred veteran — no alternatives = no decision. Only enters for missing baseline / new tech selection / architecture risk |
| 🟢 003 | [`frontend-dev`](.claude/agents/frontend-dev.md) | Components are contracts — React+Vite + Unit + SIT self-run |
| 🟢 004 | [`backend-dev`](.claude/agents/backend-dev.md) | Schema migration PTSD — FastAPI / SQLAlchemy + Unit + SIT self-run |
| 🟢 005 | [`ai-agent-dev`](.claude/agents/ai-agent-dev.md) | Prompts are code — LLM / RAG / Agent + Unit + SIT self-run |
| 🟢 006 | [`ml-engineer`](.claude/agents/ml-engineer.md) | Quality-latency-cost triangle — multimodal inference + Unit + SIT self-run |
| 🟢 007 | [`uiux-designer`](.claude/agents/uiux-designer.md) | spec.md + index.html dual delivery + AC coverage table |
| 🟢 008 | [`miniapp-dev`](.claude/agents/miniapp-dev.md) | Native-first, Taro fallback — setData 256KB is the hard wall |
| 🟡 009 | [`code-reviewer`](.claude/agents/code-reviewer.md) | Writes only to docs/reviews/ — code review + SIT Audit (4 checks + 3-tier verdict) |
| 🟡 010 | [`miniapp-code-reviewer`](.claude/agents/miniapp-code-reviewer.md) | Zero tolerance for review red lines — wx.* + bundle size + privacy + SIT Audit |
| 🔴 011 | [`qa-engineer`](.claude/agents/qa-engineer.md) | Default verdict ❌ — no curl/screenshot/DB diff = no Pass. E2E + UAT |
| 🔴 012 | [`miniapp-qa-engineer`](.claude/agents/miniapp-qa-engineer.md) | E2E without real-device QR scan = no signoff |
| 🟪 013 | [`content-writer`](.claude/agents/content-writer.md) | Numbers over adjectives, hook in first sentence — release notes / blog |
| 🟪 014 | [`growth-analyst`](.claude/agents/growth-analyst.md) | Metrics before instinct — North Star / OMTM / A/B experiments |
Capability baseline (model / tools / preloaded skills): [`team-roles.md`](.claude/standards/team-roles.md); end-to-end message flow: [`team-capability-map.md §2`](./docs/team-capability-map.md).
---
## 🔄 Process / HOW
> Full chain for a feature from requirement to delivery. **6 stage gates progress in order; no skipping on failure.**
### End-to-end overview
```mermaid
flowchart LR
U([👤 User]) --> S1["S1 Discovery
brainstorming → PRD"]
S1 --> S2["S2 Dispatch
6-section schema + AC quote"]
S2 --> POOL{"≥2 same-type
pending tasks?"}
POOL -->|yes| FANOUT["⚙️ Pool fan-out
spawn type-1 / type-2 / ...
ADR-001"]
POOL -->|no| S3
FANOUT --> S3["S3 TDD Implementation
red→green→refactor
+ Unit + SIT self-run"]
S3 --> FANIN["agf-matrix.sh
PL fan-in aggregation
(pool mode only)"]
FANIN --> S4["S4 Code Review
+ SIT Audit
3-tier verdict"]
S4 --> S5["S5 E2E
chrome-devtools-mcp
3-tier verdict"]
S5 --> S6["S6 UAT
P0 case pass^2
PL business signoff"]
S6 --> D([🎁 Delivery])
S6 -.->|optional| POST["Post-Launch
content + growth"]
S4 -.block / Redo SIT.-> S3
S5 -.Block.-> S3
S6 -.request changes.-> S3
style U fill:#3b82f6,color:#fff
style S6 fill:#f97316,color:#fff
style D fill:#22c55e,color:#fff
style FANOUT fill:#7c3aed,color:#fff
style FANIN fill:#7c3aed,color:#fff
```
Full diagram (with subgraphs + mini-program variant + post-launch chain + stage × Hook + stage × Skill + full Pool quick-reference): [`team-capability-map.md §1.1-§1.7`](./docs/team-capability-map.md).
### One feature = one Sprint
| Sprint stage | Who + what | Artifact |
|---|---|---|
| **Planning** | #001 `brainstorming` clarifies → PRD → `writing-plans` plan | `docs/prd/-.md` (10 sections / AC P0/P1) |
| **Design** | #007 spec.md + standalone HTML prototype + **AC coverage table** | `docs/design//{spec.md,index.html}` |
| **TDD Implementation** | #003 + #004 parallel (separate worktrees) → red → green → refactor; **≥2 same-type pending tasks → auto fan-out into a Pool (instances named `-N`)** | code + `tests/sit/` |
| **Async Standup** | After each Task → append `progress/{-N}.md` (5 sections: state / Skills / SIT evidence [with AC `[x]/[ ]` inline] / quality gate / next step); in Pool mode PL fans-in with `agf-matrix.sh --type=progress` | `progress/*.md` |
| **Code Review + SIT Audit** | #009 dual verdict (code `approve` / `approve with changes` / `block` + SIT Audit `Pass` / `Pass with concerns` / `Redo SIT`); Pool mode `-r` | `docs/reviews/(-r)-.md` |
| **Sprint Review (E2E + UAT)** | #011 runs E2E (verdict `Promote` / `Conditional promote` / `Block`) → UAT (**P0 case pass^2 / passes 2 times in a row** before business signoff `approve` / `request changes`); Pool mode `-e2e-q` / `-uat-q` | `docs/qa/-{e2e,uat}(-q)-.md` |
| **Signoff + archive** | #001 business signoff → `archive-progress.sh` archives progress | `docs/qa/-process-log.md` |
| **Retrospective** (MAJOR/MINOR) | #001 `/agf-release-retro vX.Y.Z` (7 steps + §5 Owner+Due mandatory) | `docs/reviews/retro-vX.Y.Z-.md` |
### PRD → Task decomposition rules
See [`product-workflow.md §3.3`](./docs/product-workflow.md#33-拆分规则硬性条款):
- 1 User Story = 1~N Tasks, **split by role boundary** (FE/BE/AI/ML), not by AC; each Task has ≥1 AC, quoted verbatim in SendMessage
- Task granularity = closed-loop within one session (implement + Unit + AC self-check)
- AC is a leaf, not splittable — if too large, split its "observable outcome"
### Failure loop (any stage fail → no skipping, return to S3)
> **4 verdict vocabularies** (team convention, no improvisation; see [`workflow.md §Verdict 词表`](.claude/standards/workflow.md)): code review `approve` / `approve with changes` / `block` · SIT Audit `Pass` / `Pass with concerns` / `Redo SIT` · QA report-level `Promote` / `Conditional promote` / `Block` · UAT business signoff `approve` / `request changes`.
| Stage gate | Failure verdict | Failure action | Returns to |
|---|---|---|---|
| **Code Review** | `block` | PL re-dispatches execution layer to fix | S3 |
| **SIT Audit** | `Redo SIT` | PL combines re-dispatch (with code fix) | S3 (does not progress to E2E) |
| **E2E** | `Block` / `Conditional promote` | qa reports → PL re-dispatches | S3 (does not progress to UAT) |
| **UAT business signoff** | `request changes` | PL re-dispatches | S3 |
| **P0 case pass^2** | fails to pass twice in a row | qa withholds signoff, returns to S3 ([`agf-writing-qa-report`](.claude/skills/agf-writing-qa-report/) enforced) | S3 |
| **Pool single-instance fail** | a single `-N` instance is blocked | PL respawns that instance (other instances keep running) | local rollback of that `-N` |
| **Pool ≥50% fail** | widespread failure | PL aborts the whole batch + PRD review | S1 |
**code-reviewer / qa-engineer never touch source directly.**
---
## ⚙️ Mechanism / ENFORCEMENT
> Agile + Scrum + TDD principles are **wired into skill / hook / DoD three-layer enforcement** — never relying on agent self-discipline.
### TDD enforcement chain
```mermaid
flowchart LR
PL([🟠 PL dispatches
new feature / bugfix]) --> HOOK["validate-task-schema
6 sections complete"]
HOOK --> SKILL["invoke skill
test-driven-development"]
SKILL --> RED["🔴 red
write failing test
commit: test:"]
RED --> GREEN["🟢 green
minimal impl turns test green
commit: feat: / fix:"]
GREEN --> REFACTOR["🔵 refactor
tidy code, keep green
commit: refactor:"]
REFACTOR --> PR["PR commit history
first test before first impl"]
PR --> CR["code-reviewer
order check"]
CR --> PASS([✅ Proceed to
code review + SIT Audit])
REFACTOR -.iterate.-> RED
CR -.violation = block.-> RED
style PL fill:#f97316,color:#fff
style PASS fill:#22c55e,color:#fff
```
**Skip condition**: pure refactor / docs / config tasks can skip TDD order ([`ac-lifecycle.md` DoD line 81–85](.claude/standards/ac-lifecycle.md)).
### 4-layer Hook defense (red lines + reminders)
| Layer | Trigger | Action |
|---|---|---|
| 🚫 | Dangerous commands: `rm -rf` / `DROP TABLE` / `git push --force` / `git reset --hard` | Hard block |
| 🚫 | Secret leaks: 10 vendor tokens + PEM/SSH/PuTTY/BIP39 mnemonics | Hard block |
| 🚫 | Prompt-injection in external content | Soft warning |
| 🚫 | commit-time staged-diff secret scan | Hard block (Edit/Write bypass backstop) |
| 💡 | Teammate wants idle but task list has pending | Reminder to stay and finish |
| 💡 | Execution layer exits without writing `progress/.md` | Block (**Daily Standup backstop**) |
| 💡 | `TaskCreate` 6-section schema incomplete | Block (**AC quote enforcement**) |
| 💡 | Pre-work skill check | Reminder (if not invoked) |
### Stage × Hook × Skill triggers (merged view)
| Stage | Hook (🛑 hard block / ⚠️ soft warning) | Skill (mandatory / recommended) |
|---|---|---|
| **prompt submit** | 🛑 `scan-secrets` (10 vendor secrets + PEM/SSH/PuTTY/BIP39)
⚠️ `skill-suggester` | — |
| **S1 discovery + PRD** | — | `brainstorming` (mandatory if vague/multi-option)
`agf-writing-prd` |
| **S2 dispatch** | 🛑 `validate-task-schema` (6 sections + AC quote) | `writing-plans` (≥3 AC / cross-role)
`using-git-worktrees` (≥2 dev parallel mandatory)
**Pool fan-out** (≥2 same-type pending tasks / [ADR-001](docs/adr/001-multi-instance-worker-pool.md)) |
| **S3 implementation tool use** | 🛑 `block-dangerous-bash`
Bash: `rm -rf` / `DROP TABLE` / `push --force` / `reset --hard`
⚠️ `sanitize-tool-output`
WebFetch/WebSearch/Read/Bash/mcp__* output prompt-injection warning | `test-driven-development` mandatory (new feature/bugfix)
`agf-running-sit-tests` (after Unit)
`systematic-debugging` (on bug)
`verification-before-completion` (before report) |
| **S3 exit/idle** | 🛑 `check-progress-file` (no progress entry → block)
🛑 `teammate-keepalive` (idle while pending → block) | — |
| **git commit** | 🛑 `scan-commit` (staged diff same secret regex) | — |
| **S4 review** | — | `requesting-code-review` (before trigger)
`receiving-code-review` (when bounced back) |
| **S5/S6 testing** | — | `agf-writing-qa-report` (E2E/UAT report) |
| **post-release**
(MAJOR/MINOR) | — | `agf-running-release-retro` (PATCH auto-aborts) |
### Release discipline + No Equivalent Bypass
**Release discipline**: MAJOR/MINOR must be tagged in the same session + tag annotation lists tagged-along commits + **must** run `/agf-release-retro`. See [`versioning.md`](.claude/standards/versioning.md).
**Hard block hit** → **No Equivalent Bypass** (must not find equivalent workaround; must SendMessage PL and wait for authorization). See [`security.md`](.claude/standards/security.md).
### 📖 Glossary
**Collaboration actions** (frequent at runtime, see [`product-workflow.md §4.5`](./docs/product-workflow.md#45-协作动作collaboration-actions)):
| Term | Meaning |
|---|---|
| **dispatch** | PL assigns Task to execution layer via SendMessage (6 sections + AC quote) |
| **fan-out / fan-in** | Pool mode: ≥2 same-type pending tasks → PL spawns `-1` / `-2` ... (fan-out) → PL aggregates the multiple reports with `agf-matrix.sh --type=progress\|review\|qa` (fan-in) |
| **pass^2** | At UAT, P0 cases must pass 2 times in a row before business signoff ([`agf-writing-qa-report`](.claude/skills/agf-writing-qa-report/) enforced) |
| **escalate** | Request upper decision (exec→PL→user; reviewer→tech-lead for architecture risk) |
| **halt / STOP ⛔** | Hook block / test fail / spec drift triggers; stop and SendMessage report — never bypass |
| **verify-before-assert** | Must actually run verification before claiming "done"; no empty assertions |
| **push back** | After review feedback, refuse with measured evidence |
| **ghost mailbox** | Dispatching to nonexistent or destroyed mailbox |
| **spec drift** | Document/PRD/spec contract diverges from actual code or database |
| **sweep** | PL actively scans git log + task list + mailbox |
| **mock-first** | Frontend implements UI with mock data, swaps to real endpoint when backend ready |
**Status indicators** (team convention, no improvisation):
| Symbol | Meaning | Typical location |
|---|---|---|
| ✅ | Done / passed | AC self-check, Code Review verdict, UAT signoff |
| ⚠️ | Warning / deviation | AC self-check has gap, Code Review Warning |
| ❌ | Failed / not passed | Test failure, Code Review request changes |
| 🔄 | In progress | Task status, long-running progress |
| ⏸ | Paused / pending | Waiting on external dep / clarification |
| 💤 | idle / standby | Teammate on standby |
| ⛔ | STOP / halt | Emergency stop (hook block / spec drift / test fail) |
---
## ⏱ Day 1
**Prerequisite**: Claude Code ≥ v2.1.154 + Agent Teams enabled (`init-team.sh` validates).
```bash
# Scaffold a new project
gh repo create my-app --template pcliangx/AppGenesisForge --private
cd my-app && bash init-team.sh
# Embed into existing repo (one-shot migration; never overwrites your CLAUDE.md / ADR-000 / settings.json)
bash install-to-existing.sh ~/path/to/your-existing-repo
# Try it
/agf-team-start
```
Full Day-1 checklist + prerequisites + pitfalls: [`docs/FIRST_RUN.md`](./docs/FIRST_RUN.md). ⚠️ Don't copy this repo's `CLAUDE.md` — the scaffold ships `CLAUDE.example.md`.
---
## 🛠️ Self-Service
### Slash Commands (5, Scrum ceremony entries)
| Command | Scrum ceremony |
|---|---|
| [`/agf-team-start`](.claude/commands/agf-team-start.md) | Sprint kickoff |
| [`/agf-uat`](.claude/commands/agf-uat.md) | Sprint Review (business signoff) |
| [`/agf-team-stop`](.claude/commands/agf-team-stop.md) | Sprint wrap-up (close execution-layer teammates after UAT) |
| [`/agf-release-retro`](.claude/commands/agf-release-retro.md) | Sprint Retrospective (PATCH auto-aborts) |
| [`/agf-tasks`](.claude/commands/agf-tasks.md) | Sprint Backlog view |
### Skills (9 project-level + 2 third-party docx/pptx + superpowers suite)
| Skill | Who / when |
|---|---|
| [`agf-writing-prd`](.claude/skills/agf-writing-prd/) | #001 writing PRD (10 sections + 5 AC rules + completion checklist) |
| [`agf-writing-adr`](.claude/skills/agf-writing-adr/) | #002 recording architecture decisions (alternatives mandatory + version verification) |
| [`agf-wiring-multi-llm-sdk`](.claude/skills/agf-wiring-multi-llm-sdk/) | #005 integrating DeepSeek/Doubao/Qwen/MiniMax |
| [`agf-running-sit-tests`](.claude/skills/agf-running-sit-tests/) | Execution-layer devs self-running SIT |
| [`agf-writing-qa-report`](.claude/skills/agf-writing-qa-report/) | #011 writing E2E/UAT reports |
| [`agf-running-release-retro`](.claude/skills/agf-running-release-retro/) | #001 running MAJOR/MINOR retros |
| [`agf-writing-docx-reports`](.claude/skills/agf-writing-docx-reports/) | Programmatic Chinese .docx reports (resolutions / reviews / RFPs) |
| [`agf-writing-pptx-reports`](.claude/skills/agf-writing-pptx-reports/) | Programmatic Chinese .pptx decks (policy / public-sector / training) |
| [`agf-writing-github-issue`](.claude/skills/agf-writing-github-issue/) | File issues in the repo (QA auto-path + locked labels) |
**Superpowers suite** (everyone can invoke): `brainstorming` / `writing-plans` / `test-driven-development` / `systematic-debugging` / `verification-before-completion` / `requesting-code-review` / `receiving-code-review` / `using-git-worktrees` — enforcement points see [`superpowers.md`](.claude/standards/superpowers.md).
### Scripts (4 PL/maintainer tools)
| Script | Who / when |
|---|---|
| [`agf-tasks.sh`](.claude/scripts/agf-tasks.sh) | PL views the task-list table (invoked internally by slash `/agf-tasks`) |
| [`agf-matrix.sh`](.claude/scripts/agf-matrix.sh) | PL Pool fan-in aggregation: `--type=progress\|review\|qa` three modes, merging N `-N.md` / `-r` / `-q` reports into one table |
| [`archive-progress.sh`](.claude/scripts/archive-progress.sh) | PL archives `progress/{-N}.md` → `docs/qa/-process-log.md` after UAT signoff (supports pool multi-instance merge) |
| [`lint-all.sh`](.claude/scripts/lint-all.sh) | Repo-wide bash/JSON/YAML lint entry; chained by the pre-commit hook |
---
## 🔍 vs Others
| | **AppGenesisForge** | Generic prompt lib | Role-persona lib | Single-agent toolchain |
|---|---|---|---|---|
| **Methodology backbone** | ✅ Agile + Scrum + TDD (Organization / Process / Mechanism three layers) | ❌ | ❌ | ❌ |
| **Organization** | ✅ 14 coworkers + Scrum role mapping | ❌ | ⚠️ role list only | ❌ |
| **Process** | ✅ PL + 6 stage gates | ❌ | ⚠️ examples only | ❌ |
| **Mechanism (TDD enforcement)** | ✅ commit order + skill + DoD | ❌ | ❌ | ❌ |
| **Mechanism (Hook defense)** | ✅ 4 layers | ❌ | ❌ | ❌ |
| **Retrospective** | ✅ 7 steps + Owner+Due mandatory | ❌ | ❌ | ❌ || **Downstream migration** | ✅ install-to-existing | ✅ copy-paste | ⚠️ fork | ⚠️ |
| **China LLMs** | ✅ DeepSeek/Doubao/Qwen/MiniMax | ❌ | ❌ | depends |
---
## ✨ Summary
- **What it is**: 14 AI coworkers + Agile + Scrum + TDD workflow + 4-layer hook defense
- **What's unique**: Agile + Scrum + TDD principles **wired into Organization + Process + Mechanism three-layer landing**, never relying on agent self-discipline
- **How to use**: `/agf-team-start ` → 14 coworkers run the Sprint flow → watch in one tab
- **Your benefit**: ship a feature overnight; artifacts are traceable; stage-gate skipping = failure of duty
> **Core trade-off**: *Wire Agile + Scrum + TDD into the enforcement path, not a wiki* — the system's most important design choice. Rules forged from real iteration: early on PL raced to write code, Reviewer out-aggressed Dev, QA delivered no evidence — every hook, standard, and skill comes from a real stumble.
### 🚀 Next Steps
| You want to | How |
|---|---|
| **Try a fresh project** | `gh repo create --template pcliangx/AppGenesisForge` → [⏱ Day 1](#-day-1) |
| **Embed into existing repo** | `bash install-to-existing.sh ~/your-repo` (never overwrites your `CLAUDE.md`/`ADR-000`/`settings.json`) |
| **Install via Plugin** (lightest) | `/plugin marketplace add pcliangx/AppGenesisForge` → `/plugin install agf` → follow [`agf-plugin/POST_INSTALL.md`](agf-plugin/POST_INSTALL.md) to add required settings + external deps |
| **See a real feature with full artifacts** | `ls docs/training/samples/postcard-feature/` — PRD → ADR → Dispatch → Progress → Review → UAT → Retro → Release notes (illustrative training paths) |
| **Give feedback / open PR** | [🐛 Issue](https://github.com/pcliangx/AppGenesisForge/issues) · [🍴 PR](https://github.com/pcliangx/AppGenesisForge/pulls) |
| **See what changed each release** | [📒 CHANGELOG](CHANGELOG.md) |
### 📂 What's Included
```text
.claude/ 14 agents · 17 standards · 11 skills · 12 hooks · 5 commands · 4 scripts · 2 rules
docs/ product-workflow / team-capability-map / PRD·QA·ADR·Review·Retro templates
docs/training/samples/ · evals/ end-to-end worked example (postcard-feature) + per-role baseline JSONL
progress/ execution-layer Self-Reporting drafts (Async Daily Standup)init-team.sh Day-1 health check (10 checks + auto pre-commit install)
install-to-existing.sh one-shot migration into existing repo
CLAUDE.example.md project-level CLAUDE.md template (don't confuse with this repo's CLAUDE.md)
```
### 🗺️ Roadmap
**Done**:
| Release | Content |
|---|---|
| Foundation | 4 China-domestic LLMs · WeChat MiniProgram trio · 4-layer hook defense · Self-Reporting Pattern · agent namespace steady state |
| **v1.6.0** | team-dashboard zero-invasion + `install-to-existing.sh` one-shot migration |
| **v1.7.0** | code-reviewer adopts `/simplify` Phase 1+2 review |
| **v2.0.0** | H3 late-stage workflow: SIT ownership dev → code-reviewer audit |
| **v3.0.0** | Removed cost audit feature (twice-empty = wrong discipline) |
| **v3.1.0** | **Agile + Scrum + TDD enforcement complete** (skill / hook / DoD three-layer) + README rewrite + docs traceability |
| **v3.1.1** | README Setup-Detail-Wrap restructure + Claude Code minimum version bumped to v2.1.130 |
| **v3.2.0** | **Multi-instance Worker Pool ([ADR-001](docs/adr/001-multi-instance-worker-pool.md))**: Dev / Reviewer / QA three-layer pool; `agf-matrix.sh` three-mode fan-in; P0 case pass^2; 4 verdict vocabularies unified |
**Up next**:
- [ ] More skills: test cases / commit messages / release notes
- [ ] More eval baselines: remaining role JSONL coverage
- [ ] CI/CD: wire `init-team.sh` into GitHub Actions
- [ ] feature cycle time trend visualization (enabled after ≥3 retros)
---
## 📜 License & Acknowledgments
MIT — free for commercial use. Structural inspiration from [The Agency](https://github.com/msitarzewski/agency-agents).
---
**14 AI coworkers · Agile + Scrum + TDD (Organization / Process / Mechanism three-layer landing) · ship a feature overnight**
[⭐ Star](https://github.com/pcliangx/AppGenesisForge) · [🍴 Fork](https://github.com/pcliangx/AppGenesisForge/fork) · [🐛 Issue](https://github.com/pcliangx/AppGenesisForge/issues) · [📒 CHANGELOG](CHANGELOG.md) · [🇨🇳 简体中文](README.md)