Picks the right tool from this collection for your request, whether planning, review, QA, shipping, docs, security or design.
garrytan/gstack
Installs Garry Tan's set of 23 opinionated Claude Code tools that act as CEO, designer, engineering manager, release manager, doc engineer and QA for your project.
- 68 skills
- 138k stars
- Reached #3 on GitHub's daily trending
The 5 most installed
Ordered by installs on skills.sh- gstack25k installs
- office-hours673 installs
Runs a YC-style office hours session to brainstorm an idea and judge whether it is worth building.
- design-review672 installs
Checks your interface for visual inconsistency, spacing and hierarchy problems and fixes them.
- plan-ceo-review561 installs
Reviews a plan in founder mode, pushing you to think bigger and expand scope.
- plan-eng-review558 installs
Locks in architecture, data flow, diagrams, edge cases, test coverage and performance before you build a feature.
The other 63 skills
- plan-design-review554 installs
Reviews a plan from a designer's perspective, focusing on the user experience.
- review524 installs
Reviews your local changes, a branch or a pull request before it lands.
- qa523 installs
Tests your site, API or CLI, finds bugs, and fixes them.
- browse481 installs
Opens web pages in a real browser, reads them, clicks through flows, takes screenshots and checks console errors.
- cso471 installs
Audits your code for security vulnerabilities and, where possible, reproduces and proposes fixes.
- ship465 installs
Merges the base, runs tests, reviews the diff, bumps version and changelog, then pushes and opens a PR.
- design-shotgun455 installs
Generates several design variants, shows them side by side and collects your feedback to iterate.
- autoplan451 installs
Runs the CEO, design, engineering and developer-experience reviews in sequence and decides each one automatically.
- qa-only447 installs
Tests your site, API or CLI and reports bugs without fixing them.
- retro434 installs
Produces a weekly engineering retrospective summarizing what your team shipped.
- document-release431 installs
Audits and updates your documentation after a release so it reflects what changed.
- careful425 installs
Warns you before the agent runs destructive commands like deleting files or resetting git.
- design-html421 installs
Turns an approved design into production-quality HTML and CSS.
- benchmark416 installs
Measures your web pages' speed and flags performance regressions against earlier runs.
- canary412 installs
Watches your site after a deploy and alerts you if something starts failing.
- health408 installs
Shows a dashboard of code quality checks for the project.
- codex399 installs
Asks OpenAI Codex CLI for a second opinion or review of your code.
- freeze398 installs
Limits the agent's file edits to one directory for the rest of the session.
- guard396 installs
Warns before destructive commands and keeps edits limited to the intended directory.
- land-and-deploy396 installs
Merges a pull request and deploys it to production.
Runs only when you type /land-and-deploy
- gstack-upgrade385 installs
Updates gstack to the latest version.
- plan-devex-review385 installs
Reviews a plan for how good the developer experience will be.
- context-save383 installs
Saves your current working context so you can resume it later.
- devex-review380 installs
Audits the developer experience of your product live, as a developer would encounter it.
- learn378 installs
Records and shows what the project has learned over time.
- context-restore377 installs
Brings back the working context you saved earlier so you can pick up where you left off.
- unfreeze371 installs
Clears the freeze boundary so all directories can be edited again.
- setup-deploy368 installs
Configures the deployment settings used by the land-and-deploy workflow.
- make-pdf360 installs
Converts a markdown file into a publication-quality PDF.
- open-gstack-browser352 installs
Launches an AI-controlled Chromium browser with the sidebar extension already installed.
- benchmark-models346 installs
Compares how different AI models such as Claude, GPT and Gemini perform on the same tasks.
- scrape342 installs
Pulls data from web pages using your real, already signed-in browser sessions.
- landing-report341 installs
Shows a read-only dashboard of the ship queue and version status.
- plan-tune335 installs
Adjusts how often gstack asks you questions, based on your preferences.
- skillify334 installs
Turns your most recent successful scrape into a reusable browser skill saved on disk.
- gstack-openclaw-office-hours332 installs
Talks through a new product idea with you and judges whether it is worth building before any code is written.
- setup-gbrain330 installs
Installs and connects gbrain for your coding agent, including a local or Supabase brain and MCP registration.
- document-generate324 installs
Writes missing documentation from scratch for a feature, module or whole project.
- gstack-openclaw-ceo-review323 installs
Reviews a plan like a CEO, challenging the proposal and deciding whether to expand or cut its scope.
- sync-gbrain321 installs
Keeps gbrain up to date with your repo's code and refreshes agent search guidance.
- gstack-openclaw-retro319 installs
Reviews a week of commits and code quality, tracking trends and each person's contributions, praise and growth areas.
- hackernews-frontpage317 installs
Scrapes the Hacker News front page, including titles, points and comment counts.
- spec312 installs
Turns a vague request into a precise, executable spec in five phases.
- ios-design-review306 installs
Audits the visual design of an iOS app running on real hardware.
- ios-qa293 installs
Tests a SwiftUI iOS app on a live device.
- ios-clean291 installs
Removes the DebugBridge package and all its debug-only wiring from an iOS app.
- ios-fix289 installs
Finds and fixes bugs in an iOS app on its own.
- ios-sync287 installs
Rebuilds the iOS debug bridge from the latest upstream gstack templates.
- diagram273 installs
Turns a text description or mermaid source into an editable Excalidraw file plus SVG and PNG images.
- test-audit54 installs
Finds low-value or duplicate tests and the test-only code they keep alive.
- eval-plan47 installs
Plans a paid AI evaluation with clear rules, dry-runs it free, then runs a small priced pilot for approval.
- design-motion42 installs
Animates an approved design into a looping MP4 and GIF showcase.
- captain33 installs
Runs several coding agents from one controller that merges their work and relays approvals.
- ubicloud33 installs
Runs a heavy test suite, build or eval on an ephemeral 16-vCPU Ubicloud virtual machine.
- modernize32 installs
Brings an old pull request onto the current base and decides to modernize, rework or close it.
Runs only when you type /modernize
- write-skill32 installs
Writes a new agent skill or improves an existing one without bloat.
- runaway-stopper23 installs
Finds runaway Capy threads and subagents, stops or rotates them safely, and prevents repeats.
- gstack-plan21 installs
Challenges a feature's scope, then reviews architecture, code quality, tests and performance and ends with a build plan.
Studies your product and its field, then proposes a full design system with font and color previews.
Connects a remote AI agent to your browser so it can work alongside you.
Runs the checks and tests until they pass before shipping a change.