gstack-workflow
Skill1.7k starsChanged 7 months agoMITView on GitHub
Structured AI-assisted development workflow with specialist roles — Think, Plan, Build, Review, Test, Ship, Reflect. Adapted from garrytan/gstack (MIT).
- Deletes or force-pushes
1 other project uses a version of this.
Trending
Quiet since February.thin dataabout the whole repository
Most recently taken up by
- DenchHQ/DenchClawword for word, first seen Oct 8, 2026
Who's talking about DenchHQ/ironclaw
The biggest 1 of 1 mention linking to the repository that our crawl found. Each line is their own headline.
Other versions1 project uses a version of this.
Word for word · 1
- DenchHQ/DenchClaw★ 1.7k100% same
Edited · 0
None found.
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository35 other files it gives its agents.
- .cursor/rules/git-rules.mdcCursor rule
- agent-builderSkill
- data-builderSkill
- game-builderSkill
- platform-apiSkill
- app-builderSkill
- composio-appsSkill
- actionsSkill
- documentsSkill
- duckdb-operationsSkill
- object-builderSkill
- reportsSkill
- database-crm-systemSkill
- views-filtersSkill
- dench-integrationsSkill
- benchmarkSkill
- canarySkill
- carefulSkill
- design-consultationSkill
- design-reviewSkill
- document-releaseSkill
- freezeSkill
- guardSkill
- investigateSkill
- land-and-deploySkill
- office-hoursSkill
- plan-ceo-reviewSkill
- plan-design-reviewSkill
- plan-eng-reviewSkill
- qa-onlySkill
- qaSkill
- retroSkill
- reviewSkill
- shipSkill
- unfreezeSkill
The file
We split this skill into the rules it gives an agent
8 rules. About 0 are common, 7 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈7
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 19
If Option A is complete (all edge cases, full coverage) and Option B is a shortcut — always recommend A.
- Distinctive: line 21
Don't skip the last 10% to "save time" — with AI, that 10% costs seconds.
- Distinctive: line 25
Before building infrastructure or unfamiliar patterns — search first.
- Distinctive: line 69
When completing any skill workflow, report status using one of:
- Distinctive: line 80
If you have attempted a task 3 times without success, STOP and escalate.
- Distinctive: line 81
If you are uncertain about a security-sensitive change, STOP and escalate.
- Distinctive: line 87
When you notice the user is at a particular stage, suggest the appropriate skill:
- Very specific: line 103
All child skills are seeded into the workspace alongside this parent skill and can be read at …
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
gstack — Structured Development Workflow
A process that turns AI into a virtual engineering team. Eighteen specialist roles, structured as a sprint: Think → Plan → Build → Review → Test → Ship → Reflect. Each skill feeds into the next.
Adapted from garrytan/gstack (MIT License).
Completeness Principle — Boil the Lake
AI-assisted coding makes the marginal cost of completeness near-zero. When presenting options:
- If Option A is complete (all edge cases, full coverage) and Option B is a shortcut — always recommend A.
- Lake vs. ocean: A "lake" is boilable — 100% test coverage for a module, full feature implementation, all edge cases. An "ocean" is not — multi-quarter migrations, rewriting dependencies. Recommend boiling lakes. Flag oceans as out of scope.
- Don't skip the last 10% to "save time" — with AI, that 10% costs seconds.
Search Before Building
Before building infrastructure or unfamiliar patterns — search first.
Three layers of knowledge:
- Layer 1 (tried and true): Don't reinvent the wheel.
- Layer 2 (new and popular): Search for these, but scrutinize.
- Layer 3 (first principles): Original observations from reasoning about the specific problem. The most valuable.
Sprint Workflow
| Phase | Skill | Role | What it does |
| Think | office-hours | YC Office Hours | Forcing questions that reframe the product before code. Design doc feeds downstream. |
| Plan | plan-ceo-review | CEO / Founder | Rethink the problem. 10-star product. Four modes: Expansion, Selective, Hold, Reduction. |
| Plan | plan-eng-review | Eng Manager | Lock architecture, data flow, diagrams, edge cases, tests. |
| Plan | plan-design-review | Senior Designer | Rate design dimensions 0-10, explain what a 10 looks like, edit the plan. |
| Plan | design-consultation | Design Partner | Build a complete design system from scratch. |
| Review | review | Staff Engineer | Find bugs that pass CI but blow up in production. Auto-fix obvious ones. |
| Review | design-review | Designer Who Codes | Same audit as plan-design-review, then fixes what it finds. |
| Debug | investigate | Debugger | Systematic root-cause debugging. No fixes without investigation. |
| Test | qa | QA Lead | Test the app, find bugs, fix with atomic commits, re-verify. |
| Test | qa-only | QA Reporter | Same methodology as qa, but report only — no code changes. |
| Test | benchmark | Performance Engineer | Page load times, Core Web Vitals, bundle sizes. Before/after on every PR. |
| Ship | ship | Release Engineer | Sync main, run tests, audit coverage, push, open PR. |
| Ship | land-and-deploy | Release Engineer | Merge PR, wait for CI, deploy, verify production health. |
| Ship | canary | SRE | Post-deploy monitoring loop. Console errors, perf regressions, page failures. |
| Reflect | retro | Eng Manager | Weekly retro with per-person breakdowns, shipping streaks, test health. |
| Reflect | document-release | Technical Writer | Update all docs to match what shipped. Catches stale READMEs. |
Safety Tools
| Skill | What it does |
| careful | Warn before destructive commands (rm -rf, DROP TABLE, force-push). |
| freeze | Restrict edits to one directory to prevent accidental changes. |
| guard | Maximum safety: careful + freeze combined. |
| unfreeze | Remove the freeze boundary. |
Completion Status Protocol
When completing any skill workflow, report status using one of:
- DONE — All steps completed successfully. Evidence provided.
- DONE_WITH_CONCERNS — Completed, but with issues to flag.
- BLOCKED — Cannot proceed. State what's blocking and what was tried.
- NEEDS_CONTEXT — Missing information required to continue.
Escalation
It is always OK to stop and say "this is too hard" or "I'm not confident in this result."
- If you have attempted a task 3 times without success, STOP and escalate.
- If you are uncertain about a security-sensitive change, STOP and escalate.
Suggesting Skills
When you notice the user is at a particular stage, suggest the appropriate skill:
- Brainstorming a new idea → suggest office-hours
- Reviewing a plan (strategy) → suggest plan-ceo-review
- Reviewing a plan (architecture) → suggest plan-eng-review
- Reviewing a plan (design) → suggest plan-design-review
- Creating a design system → suggest design-consultation
- Debugging errors → suggest investigate
- Testing the app → suggest qa
- Code review before merge → suggest review
- Visual design audit → suggest design-review
- Ready to deploy / create PR → suggest ship
- Post-ship doc updates → suggest document-release
- Weekly retrospective → suggest retro
- Working with production or live systems → suggest careful
All child skills are seeded into the workspace alongside this parent skill and can be read at {{WORKSPACE_PATH}}/skills/gstack/<child>/SKILL.md.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.