Skip to content
Back to search

gsd-pro

GSD Pro — The most powerful AI coding workflow for Claude Code. Free alternative to Cursor Composer, Copilot Workspace, Devin & Bolt. Multi-model routing, rollback/recovery, adaptive context. Fork of 30K★ original. npx get-shit-done-cc

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/itsjwill--gsd-pro/badge.svg)](/listings/itsjwill--gsd-pro)

HTML

<a href="/listings/itsjwill--gsd-pro"><img src="/api/listings/itsjwill--gsd-pro/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

npx get-shit-done-cc@latest

The installer prompts you to choose: 1. **Runtime** — Claude Code, OpenCode, Gemini, Codex, Copilot, or all 2. **Location** — Global (all projects) or local (current project only)

Verify with: - Claude Code / Gemini: `/gsd:help` - OpenCode: `/gsd-help` - Codex: `$gsd-help` - Copilot: `/gsd:help`

> [!NOTE] > Codex installation uses skills (`skills/gsd-*/SKILL.md`) rather than custom prompts.

Staying Updated

GSD evolves fast. Update periodically:

npx get-shit-done-cc@latest

<details> <summary><strong>Non-interactive Install (Docker, CI, Scripts)</strong></summary>

# Claude Code
npx get-shit-done-cc --claude --global   # Install to ~/.claude/
npx get-shit-done-cc --claude --local    # Install to ./.claude/

# OpenCode (open source, free models)
npx get-shit-done-cc --opencode --global # Install to ~/.config/opencode/

# Gemini CLI
npx get-shit-done-cc --gemini --global   # Install to ~/.gemini/

# Codex (skills-first)
npx get-shit-done-cc --codex --global    # Install to ~/.codex/
npx get-shit-done-cc --codex --local     # Install to ./.codex/

# Copilot (GitHub Copilot CLI)
npx get-shit-done-cc --copilot --global  # Install to ~/.github/
npx get-shit-done-cc --copilot --local   # Install to ./.github/

# All runtimes
npx get-shit-done-cc --all --global      # Install to all directories

Use `--global` (`-g`) or `--local` (`-l`) to skip the location prompt. Use `--claude`, `--opencode`, `--gemini`, `--codex`, `--copilot`, or `--all` to skip the runtime prompt.

</details>

<details> <summary><strong>Development Installation</strong></summary>

Clone the repository and run the installer locally:

git clone https://github.com/glittercowboy/get-shit-done.git
cd get-shit-done
node bin/install.js --claude --local

Installs to `./.claude/` for testing modifications before contributing.

</details>

Recommended: Skip Permissions Mode

GSD is designed for frictionless automation. Run Claude Code with:

claude --dangerously-skip-permissions

> [!TIP] > This is how GSD is intended to be used — stopping to approve `date` and `git commit` 50 times defeats the purpose.

<details> <summary><strong>Alternative: Granular Permissions</strong></summary>

If you prefer not to use that flag, add this to your project's `.claude/settings.json`:

{
  "permissions": {
    "allow": [
      "Bash(date:*)",
      "Bash(echo:*)",
      "Bash(cat:*)",
      "Bash(ls:*)",
      "Bash(mkdir:*)",
      "Bash(wc:*)",
      "Bash(head:*)",
      "Bash(tail:*)",
      "Bash(sort:*)",
      "Bash(grep:*)",
      "Bash(tr:*)",
      "Bash(git add:*)",
      "Bash(git commit:*)",
      "Bash(git status:*)",
      "Bash(git log:*)",
      "Bash(git diff:*)",
      "Bash(git tag:*)"
    ]
  }
}

</details>

---

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 799 lines · 1 finding marked

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. <div align="center">
  2. # GSD Pro — The Most Powerful AI Coding Workflow for Claude Code
  3. ### Free alternative to Cursor Composer, GitHub Copilot Workspace, Windsurf Cascade, Devin, and Bolt.new
  4. **English** · [简体中文](README.zh-CN.md)
  5. **A light-weight and powerful meta-prompting, context engineering and spec-driven development system for Claude Code, OpenCode, Gemini CLI, and Codex.**
  6. **Solves context rot — the quality degradation that happens as Claude fills its context window.**
  7. > **This fork** adds rollback/recovery, adaptive context loading, and model routing intelligence on top of the [original GSD](https://github.com/glittercowboy/get-shit-done) by TÂCHES. Synced with upstream v1.24.0.
  8. ![GitHub stars](https://img.shields.io/github/stars/itsjwill/gsd-pro?style=for-the-badge&logo=github&color=181717)
  9. ![Updated](https://img.shields.io/badge/Updated-2026-blue?style=for-the-badge)
  10. [![npm version](https://img.shields.io/npm/v/get-shit-done-cc?style=for-the-badge&logo=npm&logoColor=white&color=CB3837)](https://www.npmjs.com/package/get-shit-done-cc)
  11. [![npm downloads](https://img.shields.io/npm/dm/get-shit-done-cc?style=for-the-badge&logo=npm&logoColor=white&color=CB3837)](https://www.npmjs.com/package/get-shit-done-cc)
  12. [![Tests](https://img.shields.io/github/actions/workflow/status/glittercowboy/get-shit-done/test.yml?branch=main&style=for-the-badge&logo=github&label=Tests)](https://github.com/glittercowboy/get-shit-done/actions/workflows/test.yml)
  13. [![Discord](https://img.shields.io/badge/Discord-Join-5865F2?style=for-the-badge&logo=discord&logoColor=white)](https://discord.gg/gsd)
  14. [![X (Twitter)](https://img.shields.io/badge/X-@gsd__foundation-000000?style=for-the-badge&logo=x&logoColor=white)](https://x.com/gsd_foundation)
  15. [![$GSD Token](https://img.shields.io/badge/$GSD-Dexscreener-1C1C1C?style=for-the-badge&logo=data:image/svg+xml;base64,PHN2ZyB3aWR0aD0iMjQiIGhlaWdodD0iMjQiIHZpZXdCb3g9IjAgMCAyNCAyNCIgZmlsbD0ibm9uZSIgeG1sbnM9Imh0dHA6Ly93d3cudzMub3JnLzIwMDAvc3ZnIj48Y2lyY2xlIGN4PSIxMiIgY3k9IjEyIiByPSIxMCIgZmlsbD0iIzAwRkYwMCIvPjwvc3ZnPg==&logoColor=00FF00)](https://dexscreener.com/solana/dwudwjvan7bzkw9zwlbyv6kspdlvhwzrqy6ebk8xzxkv)
  16. [![License](https://img.shields.io/badge/license-MIT-blue?style=for-the-badge)](LICENSE)
  17. <br>
  18. ```bash
  19. npx get-shit-done-cc@latest
  20. ```
  21. **Works on Mac, Windows, and Linux.**
  22. <br>
  23. ![GSD Install](assets/terminal.svg)
  24. <br>
  25. *"If you know clearly what you want, this WILL build it for you. No bs."*
  26. *"I've done SpecKit, OpenSpec and Taskmaster — this has produced the best results for me."*
  27. *"By far the most powerful addition to my Claude Code. Nothing over-engineered. Literally just gets shit done."*
  28. <br>
  29. **Trusted by engineers at Amazon, Google, Shopify, and Webflow.**
  30. [Why I Built This](#why-i-built-this) · [How It Works](#how-it-works) · [Commands](#commands) · [Why It Works](#why-it-works) · [User Guide](docs/USER-GUIDE.md)
  31. </div>
  32. ---
  33. ## Why I Built This
  34. I'm a solo developer. I don't write code — Claude Code does.
  35. Other spec-driven development tools exist; BMAD, Speckit... But they all seem to make things way more complicated than they need to be (sprint ceremonies, story points, stakeholder syncs, retrospectives, Jira workflows) or lack real big picture understanding of what you're building. I'm not a 50-person software company. I don't want to play enterprise theater. I'm just a creative person trying to build great things that work.
  36. So I built GSD. The complexity is in the system, not in your workflow. Behind the scenes: context engineering, XML prompt formatting, subagent orchestration, state management. What you see: a few commands that just work.
  37. The system gives Claude everything it needs to do the work *and* verify it. I trust the workflow. It just does a good job.
  38. That's what this is. No enterprise roleplay bullshit. Just an incredibly effective system for building cool stuff consistently using Claude Code.
  39. — **TÂCHES**
  40. ---
  41. Vibecoding has a bad reputation. You describe what you want, AI generates code, and you get inconsistent garbage that falls apart at scale.
  42. GSD fixes that. It's the context engineering layer that makes Claude Code reliable. Describe your idea, let the system extract everything it needs to know, and let Claude Code get to work.
  43. ---
  44. ## Who This Is For
  45. People who want to describe what they want and have it built correctly — without pretending they're running a 50-person engineering org.
  46. ---
  47. ## Getting Started
  48. ```bash
  49. npx get-shit-done-cc@latest
  50. ```
  51. The installer prompts you to choose:
  52. 1. **Runtime** — Claude Code, OpenCode, Gemini, Codex, Copilot, or all
  53. 2. **Location** — Global (all projects) or local (current project only)
  54. Verify with:
  55. - Claude Code / Gemini: `/gsd:help`
  56. - OpenCode: `/gsd-help`
  57. - Codex: `$gsd-help`
  58. - Copilot: `/gsd:help`
  59. > [!NOTE]
  60. > Codex installation uses skills (`skills/gsd-*/SKILL.md`) rather than custom prompts.
  61. ### Staying Updated
  62. GSD evolves fast. Update periodically:
  63. ```bash
  64. npx get-shit-done-cc@latest
  65. ```
  66. <details>
  67. <summary><strong>Non-interactive Install (Docker, CI, Scripts)</strong></summary>
  68. ```bash
  69. # Claude Code
  70. npx get-shit-done-cc --claude --global # Install to ~/.claude/
  71. npx get-shit-done-cc --claude --local # Install to ./.claude/
  72. # OpenCode (open source, free models)
  73. npx get-shit-done-cc --opencode --global # Install to ~/.config/opencode/
  74. # Gemini CLI
  75. npx get-shit-done-cc --gemini --global # Install to ~/.gemini/
  76. # Codex (skills-first)
  77. npx get-shit-done-cc --codex --global # Install to ~/.codex/
  78. npx get-shit-done-cc --codex --local # Install to ./.codex/
  79. # Copilot (GitHub Copilot CLI)
  80. npx get-shit-done-cc --copilot --global # Install to ~/.github/
  81. npx get-shit-done-cc --copilot --local # Install to ./.github/
  82. # All runtimes
  83. npx get-shit-done-cc --all --global # Install to all directories
  84. ```
  85. Use `--global` (`-g`) or `--local` (`-l`) to skip the location prompt.
  86. Use `--claude`, `--opencode`, `--gemini`, `--codex`, `--copilot`, or `--all` to skip the runtime prompt.
  87. </details>
  88. <details>
  89. <summary><strong>Development Installation</strong></summary>
  90. Clone the repository and run the installer locally:
  91. ```bash
  92. git clone https://github.com/glittercowboy/get-shit-done.git
  93. cd get-shit-done
  94. node bin/install.js --claude --local
  95. ```
  96. Installs to `./.claude/` for testing modifications before contributing.
  97. </details>
  98. ### Recommended: Skip Permissions Mode
  99. GSD is designed for frictionless automation. Run Claude Code with:
  100. ```bash
  101. claude --dangerously-skip-permissions
  102. ```
  103. > [!TIP]
  104. > This is how GSD is intended to be used — stopping to approve `date` and `git commit` 50 times defeats the purpose.
  105. <details>
  106. <summary><strong>Alternative: Granular Permissions</strong></summary>
  107. If you prefer not to use that flag, add this to your project's `.claude/settings.json`:
  108. ```json
  109. {
  110. "permissions": {
  111. "allow": [
  112. "Bash(date:*)",
  113. "Bash(echo:*)",
  114. "Bash(cat:*)",
  115. "Bash(ls:*)",
  116. "Bash(mkdir:*)",
  117. "Bash(wc:*)",
  118. "Bash(head:*)",
  119. "Bash(tail:*)",
  120. "Bash(sort:*)",
  121. "Bash(grep:*)",
  122. "Bash(tr:*)",
  123. "Bash(git add:*)",
  124. "Bash(git commit:*)",
  125. "Bash(git status:*)",
  126. "Bash(git log:*)",
  127. "Bash(git diff:*)",
  128. "Bash(git tag:*)"
  129. ]
  130. }
  131. }
  132. ```
  133. </details>
  134. ---
  135. ## How It Works
  136. > **Already have code?** Run `/gsd:map-codebase` first. It spawns parallel agents to analyze your stack, architecture, conventions, and concerns. Then `/gsd:new-project` knows your codebase — questions focus on what you're adding, and planning automatically loads your patterns.
  137. ### 1. Initialize Project
  138. ```
  139. /gsd:new-project
  140. ```
  141. One command, one flow. The system:
  142. 1. **Questions** — Asks until it understands your idea completely (goals, constraints, tech preferences, edge cases)
  143. 2. **Research** — Spawns parallel agents to investigate the domain (optional but recommended)
  144. 3. **Requirements** — Extracts what's v1, v2, and out of scope
  145. 4. **Roadmap** — Creates phases mapped to requirements
  146. You approve the roadmap. Now you're ready to build.
  147. **Creates:** `PROJECT.md`, `REQUIREMENTS.md`, `ROADMAP.md`, `STATE.md`, `.planning/research/`
  148. ---
  149. ### 2. Discuss Phase
  150. ```
  151. /gsd:discuss-phase 1
  152. ```
  153. **This is where you shape the implementation.**
  154. Your roadmap has a sentence or two per phase. That's not enough context to build something the way *you* imagine it. This step captures your preferences before anything gets researched or planned.
  155. The system analyzes the phase and identifies gray areas based on what's being built:
  156. - **Visual features** → Layout, density, interactions, empty states
  157. - **APIs/CLIs** → Response format, flags, error handling, verbosity
  158. - **Content systems** → Structure, tone, depth, flow
  159. - **Organization tasks** → Grouping criteria, naming, duplicates, exceptions
  160. For each area you select, it asks until you're satisfied. The output — `CONTEXT.md` — feeds directly into the next two steps:
  161. 1. **Researcher reads it** — Knows what patterns to investigate ("user wants card layout" → research card component libraries)
  162. 2. **Planner reads it** — Knows what decisions are locked ("infinite scroll decided" → plan includes scroll handling)
  163. The deeper you go here, the more the system builds what you actually want. Skip it and you get reasonable defaults. Use it and you get *your* vision.
  164. **Creates:** `{phase_num}-CONTEXT.md`
  165. ---
  166. ### 3. Plan Phase
  167. ```
  168. /gsd:plan-phase 1
  169. ```
  170. The system:
  171. 1. **Researches** — Investigates how to implement this phase, guided by your CONTEXT.md decisions
  172. 2. **Plans** — Creates 2-3 atomic task plans with XML structure
  173. 3. **Verifies** — Checks plans against requirements, loops until they pass
  174. Each plan is small enough to execute in a fresh context window. No degradation, no "I'll be more concise now."
  175. **Creates:** `{phase_num}-RESEARCH.md`, `{phase_num}-{N}-PLAN.md`
  176. ---
  177. ### 4. Execute Phase
  178. ```
  179. /gsd:execute-phase 1
  180. ```
  181. The system:
  182. 1. **Runs plans in waves** — Parallel where possible, sequential when dependent
  183. 2. **Fresh context per plan** — 200k tokens purely for implementation, zero accumulated garbage
  184. 3. **Commits per task** — Every task gets its own atomic commit
  185. 4. **Verifies against goals** — Checks the codebase delivers what the phase promised
  186. Walk away, come back to completed work with clean git history.
  187. **How Wave Execution Works:**
  188. Plans are grouped into "waves" based on dependencies. Within each wave, plans run in parallel. Waves run sequentially.
  189. ```
  190. ┌────────────────────────────────────────────────────────────────────┐
  191. │ PHASE EXECUTION │
  192. ├────────────────────────────────────────────────────────────────────┤
  193. │ │
  194. │ WAVE 1 (parallel) WAVE 2 (parallel) WAVE 3 │
  195. │ ┌─────────┐ ┌─────────┐ ┌─────────┐ ┌─────────┐ ┌─────────┐ │
  196. │ │ Plan 01 │ │ Plan 02 │ → │ Plan 03 │ │ Plan 04 │ → │ Plan 05 │ │
  197. │ │ │ │ │ │ │ │ │ │ │ │
  198. │ │ User │ │ Product │ │ Orders │ │ Cart │ │ Checkout│ │
  199. │ │ Model │ │ Model │ │ API │ │ API │ │ UI │ │
  200. │ └─────────┘ └─────────┘ └─────────┘ └─────────┘ └─────────┘ │
  201. │ │ │ ↑ ↑ ↑ │
  202. │ └───────────┴──────────────┴───────────┘ │ │
  203. │ Dependencies: Plan 03 needs Plan 01 │ │
  204. │ Plan 04 needs Plan 02 │ │
  205. │ Plan 05 needs Plans 03 + 04 │ │
  206. │ │
  207. └────────────────────────────────────────────────────────────────────┘
  208. ```
  209. **Why waves matter:**
  210. - Independent plans → Same wave → Run in parallel
  211. - Dependent plans → Later wave → Wait for dependencies
  212. - File conflicts → Sequential plans or same plan
  213. This is why "vertical slices" (Plan 01: User feature end-to-end) parallelize better than "horizontal layers" (Plan 01: All models, Plan 02: All APIs).
  214. **Creates:** `{phase_num}-{N}-SUMMARY.md`, `{phase_num}-VERIFICATION.md`
  215. ---
  216. ### 5. Verify Work
  217. ```
  218. /gsd:verify-work 1
  219. ```
  220. **This is where you confirm it actually works.**
  221. Automated verification checks that code exists and tests pass. But does the feature *work* the way you expected? This is your chance to use it.
  222. The system:
  223. 1. **Extracts testable deliverables** — What you should be able to do now
  224. 2. **Walks you through one at a time** — "Can you log in with email?" Yes/no, or describe what's wrong
  225. 3. **Diagnoses failures automatically** — Spawns debug agents to find root causes
  226. 4. **Creates verified fix plans** — Ready for immediate re-execution
  227. If everything passes, you move on. If something's broken, you don't manually debug — you just run `/gsd:execute-phase` again with the fix plans it created.
  228. **Creates:** `{phase_num}-UAT.md`, fix plans if issues found
  229. ---
  230. ### 6. Repeat → Complete → Next Milestone
  231. ```
  232. /gsd:discuss-phase 2
  233. /gsd:plan-phase 2
  234. /gsd:execute-phase 2
  235. /gsd:verify-work 2
  236. ...
  237. /gsd:complete-milestone
  238. /gsd:new-milestone
  239. ```
  240. Loop **discuss → plan → execute → verify** until milestone complete.
  241. If you want faster intake during discussion, use `/gsd:discuss-phase <n> --batch` to answer a small grouped set of questions at once instead of one-by-one.
  242. Each phase gets your input (discuss), proper research (plan), clean execution (execute), and human verification (verify). Context stays fresh. Quality stays high.
  243. When all phases are done, `/gsd:complete-milestone` archives the milestone and tags the release.
  244. Then `/gsd:new-milestone` starts the next version — same flow as `new-project` but for your existing codebase. You describe what you want to build next, the system researches the domain, you scope requirements, and it creates a fresh roadmap. Each milestone is a clean cycle: define → build → ship.
  245. ---
  246. ### Quick Mode
  247. ```
  248. /gsd:quick
  249. ```
  250. **For ad-hoc tasks that don't need full planning.**
  251. Quick mode gives you GSD guarantees (atomic commits, state tracking) with a faster path:
  252. - **Same agents** — Planner + executor, same quality
  253. - **Skips optional steps** — No research, no plan checker, no verifier by default
  254. - **Separate tracking** — Lives in `.planning/quick/`, not phases
  255. **`--discuss` flag:** Lightweight discussion to surface gray areas before planning.
  256. **`--research` flag:** Spawns a focused researcher before planning. Investigates implementation approaches, library options, and pitfalls. Use when you're unsure how to approach a task.
  257. **`--full` flag:** Enables plan-checking (max 2 iterations) and post-execution verification.
  258. Flags are composable: `--discuss --research --full` gives discussion + research + plan-checking + verification.
  259. ```
  260. /gsd:quick
  261. > What do you want to do? "Add dark mode toggle to settings"
  262. ```
  263. **Creates:** `.planning/quick/001-add-dark-mode-toggle/PLAN.md`, `SUMMARY.md`
  264. ---
  265. ## Why It Works
  266. ### Context Engineering
  267. Claude Code is incredibly powerful *if* you give it the context it needs. Most people don't.
  268. GSD handles it for you:
  269. | File | What it does |
  270. |------|--------------|
  271. | `PROJECT.md` | Project vision, always loaded |
  272. | `research/` | Ecosystem knowledge (stack, features, architecture, pitfalls) |
  273. | `REQUIREMENTS.md` | Scoped v1/v2 requirements with phase traceability |
  274. | `ROADMAP.md` | Where you're going, what's done |
  275. | `STATE.md` | Decisions, blockers, position — memory across sessions |
  276. | `PLAN.md` | Atomic task with XML structure, verification steps |
  277. | `SUMMARY.md` | What happened, what changed, committed to history |
  278. | `todos/` | Captured ideas and tasks for later work |
  279. Size limits based on where Claude's quality degrades. Stay under, get consistent excellence.
  280. ### XML Prompt Formatting
  281. Every plan is structured XML optimized for Claude:
  282. ```xml
  283. <task type="auto">
  284. <name>Create login endpoint</name>
  285. <files>src/app/api/auth/login/route.ts</files>
  286. <action>
  287. Use jose for JWT (not jsonwebtoken - CommonJS issues).
  288. Validate credentials against users table.
  289. Return httpOnly cookie on success.
  290. </action>
  291. <verify>curl -X POST localhost:3000/api/auth/login returns 200 + Set-Cookie</verify>
  292. <done>Valid credentials return cookie, invalid return 401</done>
  293. </task>
  294. ```
  295. Precise instructions. No guessing. Verification built in.
  296. ### Multi-Agent Orchestration
  297. Every stage uses the same pattern: a thin orchestrator spawns specialized agents, collects results, and routes to the next step.
  298. | Stage | Orchestrator does | Agents do |
  299. |-------|------------------|-----------|
  300. | Research | Coordinates, presents findings | 4 parallel researchers investigate stack, features, architecture, pitfalls |
  301. | Planning | Validates, manages iteration | Planner creates plans, checker verifies, loop until pass |
  302. | Execution | Groups into waves, tracks progress | Executors implement in parallel, each with fresh 200k context |
  303. | Verification | Presents results, routes next | Verifier checks codebase against goals, debuggers diagnose failures |
  304. The orchestrator never does heavy lifting. It spawns agents, waits, integrates results.
  305. **The result:** You can run an entire phase — deep research, multiple plans created and verified, thousands of lines of code written across parallel executors, automated verification against goals — and your main context window stays at 30-40%. The work happens in fresh subagent contexts. Your session stays fast and responsive.
  306. ### Atomic Git Commits
  307. Each task gets its own commit immediately after completion:
  308. ```bash
  309. abc123f docs(08-02): complete user registration plan
  310. def456g feat(08-02): add email confirmation flow
  311. hij789k feat(08-02): implement password hashing
  312. lmn012o feat(08-02): create registration endpoint
  313. ```
  314. > [!NOTE]
  315. > **Benefits:** Git bisect finds exact failing task. Each task independently revertable. Clear history for Claude in future sessions. Better observability in AI-automated workflow.
  316. Every commit is surgical, traceable, and meaningful.
  317. ### Modular by Design
  318. - Add phases to current milestone
  319. - Insert urgent work between phases
  320. - Complete milestones and start fresh
  321. - Adjust plans without rebuilding everything
  322. You're never locked in. The system adapts.
  323. ---
  324. ## Commands
  325. ### Core Workflow
  326. | Command | What it does |
  327. |---------|--------------|
  328. | `/gsd:new-project [--auto]` | Full initialization: questions → research → requirements → roadmap |
  329. | `/gsd:discuss-phase [N] [--auto]` | Capture implementation decisions before planning |
  330. | `/gsd:plan-phase [N] [--auto]` | Research + plan + verify for a phase |
  331. | `/gsd:execute-phase <N>` | Execute all plans in parallel waves, verify when complete |
  332. | `/gsd:verify-work [N]` | Manual user acceptance testing ¹ |
  333. | `/gsd:audit-milestone` | Verify milestone achieved its definition of done |
  334. | `/gsd:complete-milestone` | Archive milestone, tag release |
  335. | `/gsd:new-milestone [name]` | Start next version: questions → research → requirements → roadmap |
  336. ### UI Design
  337. | Command | What it does |
  338. |---------|--------------|
  339. | `/gsd:ui-phase [N]` | Generate UI design contract (UI-SPEC.md) for frontend phases |
  340. | `/gsd:ui-review [N]` | Retroactive 6-pillar visual audit of implemented frontend code |
  341. ### Navigation
  342. | Command | What it does |
  343. |---------|--------------|
  344. | `/gsd:progress` | Where am I? What's next? |
  345. | `/gsd:help` | Show all commands and usage guide |
  346. | `/gsd:update` | Update GSD with changelog preview |
  347. | `/gsd:join-discord` | Join the GSD Discord community |
  348. ### Brownfield
  349. | Command | What it does |
  350. |---------|--------------|
  351. | `/gsd:map-codebase [area]` | Analyze existing codebase before new-project |
  352. ### Phase Management
  353. | Command | What it does |
  354. |---------|--------------|
  355. | `/gsd:add-phase` | Append phase to roadmap |
  356. | `/gsd:insert-phase [N]` | Insert urgent work between phases |
  357. | `/gsd:remove-phase [N]` | Remove future phase, renumber |
  358. | `/gsd:list-phase-assumptions [N]` | See Claude's intended approach before planning |
  359. | `/gsd:plan-milestone-gaps` | Create phases to close gaps from audit |
  360. ### Session
  361. | Command | What it does |
  362. |---------|--------------|
  363. | `/gsd:pause-work` | Create handoff when stopping mid-phase |
  364. | `/gsd:resume-work` | Restore from last session |
  365. ### Utilities
  366. | Command | What it does |
  367. |---------|--------------|
  368. | `/gsd:settings` | Configure model profile and workflow agents |
  369. | `/gsd:set-profile <profile>` | Switch model profile (quality/balanced/budget/inherit) |
  370. | `/gsd:add-todo [desc]` | Capture idea for later |
  371. | `/gsd:check-todos` | List pending todos |
  372. | `/gsd:debug [desc]` | Systematic debugging with persistent state |
  373. | `/gsd:quick [--full] [--discuss] [--research]` | Execute ad-hoc task with GSD guarantees (`--full` adds plan-checking and verification, `--discuss` gathers context first, `--research` investigates approaches before planning) |
  374. | `/gsd:health [--repair]` | Validate `.planning/` directory integrity, auto-repair with `--repair` |
  375. | `/gsd:stats` | Display project statistics — phases, plans, requirements, git metrics |
  376. <sup>¹ Contributed by reddit user OracleGreyBeard</sup>
  377. ---
  378. ## Configuration
  379. GSD stores project settings in `.planning/config.json`. Configure during `/gsd:new-project` or update later with `/gsd:settings`. For the full config schema, workflow toggles, git branching options, and per-agent model breakdown, see the [User Guide](docs/USER-GUIDE.md#configuration-reference).
  380. ### Core Settings
  381. | Setting | Options | Default | What it controls |
  382. |---------|---------|---------|------------------|
  383. | `mode` | `yolo`, `interactive` | `interactive` | Auto-approve vs confirm at each step |
  384. | `granularity` | `coarse`, `standard`, `fine` | `standard` | Phase granularity — how finely scope is sliced (phases × plans) |
  385. ### Model Profiles
  386. Control which Claude model each agent uses. Balance quality vs token spend.
  387. | Profile | Planning | Execution | Verification |
  388. |---------|----------|-----------|--------------|
  389. | `quality` | Opus | Opus | Sonnet |
  390. | `balanced` (default) | Opus | Sonnet | Sonnet |
  391. | `budget` | Sonnet | Sonnet | Haiku |
  392. | `inherit` | Inherit | Inherit | Inherit |
  393. Switch profiles:
  394. ```
  395. /gsd:set-profile budget
  396. ```
  397. Use `inherit` to follow the current runtime model selection (for example OpenCode `/model`).
  398. Or configure via `/gsd:settings`.
  399. ### Workflow Agents
  400. These spawn additional agents during planning/execution. They improve quality but add tokens and time.
  401. | Setting | Default | What it does |
  402. |---------|---------|--------------|
  403. | `workflow.research` | `true` | Researches domain before planning each phase |
  404. | `workflow.plan_check` | `true` | Verifies plans achieve phase goals before execution |
  405. | `workflow.verifier` | `true` | Confirms must-haves were delivered after execution |
  406. | `workflow.auto_advance` | `false` | Auto-chain discuss → plan → execute without stopping |
  407. Use `/gsd:settings` to toggle these, or override per-invocation:
  408. - `/gsd:plan-phase --skip-research`
  409. - `/gsd:plan-phase --skip-verify`
  410. ### Execution
  411. | Setting | Default | What it controls |
  412. |---------|---------|------------------|
  413. | `parallelization.enabled` | `true` | Run independent plans simultaneously |
  414. | `planning.commit_docs` | `true` | Track `.planning/` in git |
  415. ### Git Branching
  416. Control how GSD handles branches during execution.
  417. | Setting | Options | Default | What it does |
  418. |---------|---------|---------|--------------|
  419. | `git.branching_strategy` | `none`, `phase`, `milestone` | `none` | Branch creation strategy |
  420. | `git.phase_branch_template` | string | `gsd/phase-{phase}-{slug}` | Template for phase branches |
  421. | `git.milestone_branch_template` | string | `gsd/{milestone}-{slug}` | Template for milestone branches |
  422. **Strategies:**
  423. - **`none`** — Commits to current branch (default GSD behavior)
  424. - **`phase`** — Creates a branch per phase, merges at phase completion
  425. - **`milestone`** — Creates one branch for entire milestone, merges at completion
  426. At milestone completion, GSD offers squash merge (recommended) or merge with history.
  427. ---
  428. ## Security
  429. ### Protecting Sensitive Files
  430. GSD's codebase mapping and analysis commands read files to understand your project. **Protect files containing secrets** by adding them to Claude Code's deny list:
  431. 1. Open Claude Code settings (`.claude/settings.json` or global)
  432. 2. Add sensitive file patterns to the deny list:
  433. ```json
  434. {
  435. "permissions": {
  436. "deny": [
  437. "Read(.env)",
  438. "Read(.env.*)",
  439. "Read(**/secrets/*)",
  440. "Read(**/*credential*)",
  441. "Read(**/*.pem)",
  442. "Read(**/*.key)"
  443. ]
  444. }
  445. }
  446. ```
  447. This prevents Claude from reading these files entirely, regardless of what commands you run.
  448. > [!IMPORTANT]
  449. > GSD includes built-in protections against committing secrets, but defense-in-depth is best practice. Deny read access to sensitive files as a first line of defense.
  450. ---
  451. ## Troubleshooting
  452. **Commands not found after install?**
  453. - Restart your runtime to reload commands/skills
  454. - Verify files exist in `~/.claude/commands/gsd/` (global) or `./.claude/commands/gsd/` (local)
  455. - For Codex, verify skills exist in `~/.codex/skills/gsd-*/SKILL.md` (global) or `./.codex/skills/gsd-*/SKILL.md` (local)
  456. **Commands not working as expected?**
  457. - Run `/gsd:help` to verify installation
  458. - Re-run `npx get-shit-done-cc` to reinstall
  459. **Updating to the latest version?**
  460. ```bash
  461. npx get-shit-done-cc@latest
  462. ```
  463. **Using Docker or containerized environments?**
  464. If file reads fail with tilde paths (`~/.claude/...`), set `CLAUDE_CONFIG_DIR` before installing:
  465. ```bash
  466. CLAUDE_CONFIG_DIR=/home/youruser/.claude npx get-shit-done-cc --global
  467. ```
  468. This ensures absolute paths are used instead of `~` which may not expand correctly in containers.
  469. ### Uninstalling
  470. To remove GSD completely:
  471. ```bash
  472. # Global installs
  473. npx get-shit-done-cc --claude --global --uninstall
  474. npx get-shit-done-cc --opencode --global --uninstall
  475. npx get-shit-done-cc --codex --global --uninstall
  476. npx get-shit-done-cc --copilot --global --uninstall
  477. # Local installs (current project)
  478. npx get-shit-done-cc --claude --local --uninstall
  479. npx get-shit-done-cc --opencode --local --uninstall
  480. npx get-shit-done-cc --codex --local --uninstall
  481. npx get-shit-done-cc --copilot --local --uninstall
  482. ```
  483. This removes all GSD commands, agents, hooks, and settings while preserving your other configurations.
  484. ---
  485. ## Community Ports
  486. OpenCode, Gemini CLI, and Codex are now natively supported via `npx get-shit-done-cc`.
  487. These community ports pioneered multi-runtime support:
  488. | Project | Platform | Description |
  489. |---------|----------|-------------|
  490. | [gsd-opencode](https://github.com/rokicool/gsd-opencode) | OpenCode | Original OpenCode adaptation |
  491. | gsd-gemini (archived) | Gemini CLI | Original Gemini adaptation by uberfuzzy |
  492. ---
  493. ## 2.0 Fork Enhancements
  494. This fork adds features not yet in upstream:
  495. ### Rollback & Recovery
  496. Safety net for experimentation — auto-created git checkpoints with one-command undo.
  497. ```
  498. /gsd:rollback last # Undo last plan execution
  499. /gsd:rollback 02-01 # Rollback to specific checkpoint
  500. /gsd:recover # Diagnose and fix interrupted states
  501. /gsd:resume-task <id> # Resume interrupted subagent
  502. ```
  503. ### Adaptive Context Loading
  504. 5-tier context system that loads only what's needed ([details](get-shit-done/core/context-loader.md)):
  505. - **Tier 0** (~200 tokens): Position + config only
  506. - **Tier 1** (~2K tokens): Planning context
  507. - **Tier 2** (~3K tokens): Execution context + dependency graph
  508. - **Tier 3** (~5K tokens): Brownfield codebase docs
  509. - **Tier 4** (~10K+ tokens): Full context for milestones
  510. Includes subsystem auto-detection (auth, database, API, UI keywords) and frontmatter dependency graphs.
  511. ### Multi-Model Routing Reference
  512. Routing guide for optimal cost/quality balance ([details](get-shit-done/core/model-router.md)):
  513. - **Haiku**: File checks, validation, progress (92% cost savings)
  514. - **Sonnet**: Code generation, planning (default)
  515. - **Opus**: Architecture, research, complex debugging
  516. ---
  517. ## Cost Comparison
  518. Why pay $20-500/month for AI coding assistants when GSD Pro gives you more control for free?
  519. | Feature | GSD Pro | Cursor ($20/mo) | Copilot Workspace ($19/mo) | Devin ($500/mo) | Bolt.new ($20/mo) |
  520. |---------|---------|-----------------|---------------------------|-----------------|-------------------|
  521. | **Context engineering** | 5-tier adaptive | Basic | Basic | Proprietary | None |
  522. | **Rollback/recovery** | One-command undo | Git only | Git only | No | No |
  523. | **Multi-model routing** | Haiku/Sonnet/Opus | GPT-4o only | GPT-4o/Claude | Claude only | Claude only |
  524. | **Subagent isolation** | Fresh 200K per task | Shared context | Shared context | Shared context | Shared context |
  525. | **Brownfield support** | 7-doc codebase map | File search | File search | Repo analysis | No |
  526. | **Parallel execution** | Multi-agent waves | Sequential | Sequential | Sequential | Sequential |
  527. | **XML task structure** | Claude-native | Prompt-based | Prompt-based | Prompt-based | Prompt-based |
  528. | **Atomic git commits** | Per-task commits | Manual | Auto (messy) | Auto (messy) | No git |
  529. | **Open source** | MIT | No | No | No | No |
  530. | **Price** | **Free** | $20/mo | $19/mo | $500/mo | $20/mo |
  531. > GSD Pro runs on top of Claude Code ($20/mo for Claude Max or pay-per-token via API). The framework itself is free and open source.
  532. ---
  533. ## More Open Source AI Tools
  534. | Repo | What It Does |
  535. |------|-------------|
  536. | [RECON](https://github.com/itsjwill/RECON) | Auto-research MCP server — deep dives on any topic with YouTube, papers, and NotebookLM |
  537. ---
  538. ## Star History
  539. <a href="https://star-history.com/#glittercowboy/get-shit-done&Date">
  540. <picture>
  541. <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/svg?repos=glittercowboy/get-shit-done&type=Date&theme=dark" />
  542. <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/svg?repos=glittercowboy/get-shit-done&type=Date" />
  543. <img alt="Star History Chart" src="https://api.star-history.com/svg?repos=glittercowboy/get-shit-done&type=Date" />
  544. </picture>
  545. </a>
  546. ---
  547. ## License
  548. MIT License. See [LICENSE](LICENSE) for details.
  549. ---
  550. <div align="center">
  551. **Claude Code is powerful. GSD makes it reliable.**
  552. </div>

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.