start
Skill4 starsChanged 46 days agoMITView on GitHub
First-time onboarding for a new web project — asks where you are, configures the stack, and routes to product-spec or game-concept. Use when starting from scratch or when technical-preferences.md is still [TO BE CONFIGURED].
130 other projects use a version of this.
Adopted by
Most recently taken up by
- ansvisor/ansvisortheir own version, first seen Nov 8, 2026
- dsifry/metaswarmtheir own version, first seen Nov 8, 2026
- ai-search-guru/getcito-worlds-first-open-source-aio-aeo-or-geo-tooltheir own version, first seen Nov 8, 2026
- paulocsb/worldcup-fifa-pooltheir own version, first seen Nov 8, 2026
- srinator22/New-Projecttheir own version, first seen Nov 8, 2026
- Memolok/Claude-plugintheir own version, first seen Nov 6, 2026
Used with
Projects with it also use setup-engine, team-audio and 8 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
CoralGameSkillsetup-engine
Configure the project's game engine and version. Pins the engine in CLAUDE.md, detects knowledge gaps, and populates engine reference docs via WebSearch when the version is beyond the LLM's training data.
Also in 13 projects that use this
CoralGameSkillteam-audio
Orchestrate audio team: audio-director + sound-designer + technical-artist + gameplay-programmer for full audio pipeline from direction to implementation.
Also in 13 projects that use this
lilangMaxSkillasset-spec
Generate per-asset visual specifications and AI generation prompts from GDDs, level docs, or character profiles. Produces structured spec files and updates the master asset manifest. Run after art bible and GDD/level design are approved, before production begins.
Also in 10 projects that use this
lilangMaxSkillcreate-control-manifest
After architecture is complete, produces a flat actionable rules sheet for programmers — what you must do, what you must never do, per system and per layer. Extracted from all Accepted ADRs, technical preferences, and engine reference docs. More immediately actionable than ADRs (which explain why).
Also in 10 projects that use this
- +5 more from CoralGame
- +1 more from lilangMax
Other versions130 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 130
- anthropics/knowledge-work-plugins★ 35ksame name
- Donchitos/Claude-Code-Game-Studios★ 27ksame name
- openai/plugins★ 7.8ksame name
- SterlingChin/marvin-template★ 1ksame name
- educlopez/ui-craft★ 441same name
- ai-search-guru/getcito-worlds-first-open-source-aio-aeo-or-geo-tool★ 432same name
- dsifry/metaswarm★ 426same name
- pixel-cellar/Claude-Code-Game-Studios★ 324same name
- ansvisor/ansvisor★ 233same name
- alexeykrol/claude-code-starter★ 194same name
120 more not listed
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 repository52 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- testing/CLAUDE.mdCLAUDE.md
- a11y-auditSkill
- adoptSkill
- api-contractSkill
- architecture-decisionSkill
- architecture-reviewSkill
- backlogSkill
- brainstormSkill
- changelogSkill
- code-reviewSkill
- collaboration-protocolSkill
- create-storiesSkill
- data-modelSkill
- dependency-auditSkill
- deploySkill
- design-systemSkill
- dev-storySkill
- docsSkill
- feature-specSkill
- game-conceptSkill
- hardenSkill
- helpSkill
- hotfixSkill
- impactSkill
- incidentSkill
- initSkill
- migrateSkill
- pentestSkill
- perf-auditSkill
- product-specSkill
- qa-planSkill
- refactorSkill
- release-checklistSkill
- retrospectiveSkill
- security-auditSkill
- setup-stackSkill
- skill-improveSkill
- skill-testSkill
- sprint-planSkill
- sprint-statusSkill
- stack-updateSkill
- story-doneSkill
- team-featureSkill
- team-gameSkill
- team-releaseSkill
- team-securitySkill
- tech-debtSkill
- test-setupSkill
- threat-modelSkill
- updateSkill
- ux-specSkill
The file
We split this skill into the rules it gives an agent
4 rules. 0 are common, 0 are distinctive, and 4 are very specific to this project.
0
Common
Most of the 117 edited versions we read keep these.
0
Distinctive
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 17
Read (without showing): .claude/docs/technical-preferences.md (exists?
- Very specific: line 38
Write the files that are absent (production/review-mode.txt, production/stage.txt = discovery) after "May I write?" — one AskUserQuestion: …
- Very specific: line 48
When HEAD is a story branch, name it and ask one AskUserQuestion: switch to the default branch and commit there (Recommended — a …
- Very specific: line 50
Before asking, record the gate — <hooks>session-state.sh set Gate "/start Phase 5: commit?" — and clear it after the answer …
“Common” is measured: kept by at least 6 in 10 of the 117 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Start
Reply in the project conversation language (CLAUDE.md → Language); code, identifiers, paths and commit messages stay in English.
Entry point for a new project. Assumes nothing — asks, then routes. Writes files only after "May I write?" → "yes".
Bash is used for two things: touch .claude/.write-consent after the Phase 3 write answer, and the Phase 5 commit with its gate record; every other step reads and asks. A command in a hand-off is /web-studio:<command> in plugin mode and /<command> in copy mode; <hooks> is .claude/hooks/ in copy mode and ${CLAUDE_PLUGIN_ROOT}/hooks/ in plugin mode.
Phase 1: Silent state detection
- Read (without showing):
.claude/docs/technical-preferences.md(exists? configured?),docs/specs/product-spec.md,docs/specs/game-concept.md, code presence (go.mod,composer.json,package.json,angular.json,nuxt.config.*),production/roadmap.md,production/review-mode.txt,production/stage.txt,CLAUDE.mdLanguage section. .claude/docs/missing → the studio is not initialised: run/web-studio:initfirst (copy mode/init; in plugin mode the bare/initis Claude Code's built-in, which does not scaffold the studio).- Code or specs already exist → stop and hand off to
/adopt. The test is mechanical, never a feeling about the repository:- specs exist when
docs/specs/holds at least one.mdfile (product-spec.md,game-concept.mdorfeatures/*.md); - code exists when a manifest (
go.mod,composer.json,package.json) sits next to a source tree — a.gofile outsidevendor/, asrc/,app/,cmd/orinternal/directory with files in it, orangular.json/nuxt.config.*. A manifest alone (an empty scaffold) is not code. Either holds → say what was found (the paths), verdictBLOCKED (existing project — run /adopt full), write nothing, and end with oneAskUserQuestion:/adopt full(Recommended — the stack and artefacts are audited andtechnical-preferences.mdis filled from the facts) ·/help· stop here./startnever re-asks the review mode or rewrites the stage of a project that already has work in it.
- specs exist when
Phase 2: Where are you
AskUserQuestion: "Where are we starting from?"
- A) Just an idea — a theme or pain, nothing written →
/brainstorm. - B) Clear product — we know what we build →
/setup-stack→/product-spec. - C) Browser game — a game concept →
/setup-stack(type game) →/game-concept→/product-spec(light). - D) Existing work — code/documents already exist →
/adopt.
Phase 3: Project type and review mode
Second question: type (site | spa | api | fullstack | game | game+backend) and, only when production/review-mode.txt is absent, the review mode
(full — all gates; lean — lead + security on sensitive work (default for solo); solo).
/init normally writes the review mode already: when the file exists, print its value in one line ("review mode: lean, set by /init") and keep it — the question is not asked again and the file is not rewritten. Changing the mode is a deliberate edit of that file, never a side effect of /start.
The same holds for production/stage.txt: absent → discovery; present with discovery → nothing to write; present with any other value → say so and leave it (a project past discovery is /adopt's case, Phase 1 step 3).
Write the files that are absent (production/review-mode.txt, production/stage.txt = discovery) after "May I write?" — one AskUserQuestion: write (Recommended) · show the draft/diff first · not now. After the "write" answer: touch .claude/.write-consent (rule 7 — the consent-guard hook checks the marker). Nothing absent → no write gate in this phase.
Phase 4: Route
- Show the next 3 steps from
.claude/docs/workflow-catalog.yamlwith commands. production/roadmap.mdmissing → offer to create it (checkbox list) or, if an external advisor skill is installed, suggest its init command..claude/settings.web-studio.jsonexists (settings.json pre-dated the install) → offer to merge hooks/permissions (show the diff, ask).
Phase 5: Commit (documents lane)
Right after the last write, one commit gate (rule 7, .claude/docs/git-workflow.md § Documents): docs: initialise web studio, staging exactly the documents this run wrote (production/review-mode.txt, production/stage.txt, production/roadmap.md when Phase 4 created it).
- On the default branch when no story work is in progress.
- When HEAD is a story branch, name it and ask one
AskUserQuestion: switch to the default branch and commit there (Recommended — a pipeline-wide document) · commit here (the document belongs to this story) · leave uncommitted. - A merged
.claude/settings.json(Phase 4 step 3) is toolchain work, not a document: it does not ride thedocs:commit. Name it in the result and offer the chore lane for it (git-workflow.md § Chore / infra). - Before asking, record the gate —
<hooks>session-state.sh set Gate "/start Phase 5: commit?"— and clear it after the answer (<hooks>session-state.sh set Gate "—").
Nothing is committed without the answer; when this run wrote nothing, there is no commit gate.
Verdict: READY — project type and review mode chosen, next step named | BLOCKED (existing project — run /adopt full). Next step — one AskUserQuestion, the route from Phase 2 first (Recommended): /setup-stack (B, C) · /brainstorm (A, or the idea is still vague) · /adopt (D) · stop here.
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.

