スキルを探す

Attach a self-contained Excalidraw status diagram to a GitHub epic — read the epic + its sub-issue tree, render boxes-coloured-by-status with dependency arrows to a committed PNG that (a) renders on the GitHub mobile app and (b) has the editable Excalidraw scene embedded inside it, post it as a sticky comment, and round-trip human edits back in from a comment. Use when asked to "diagram an epic", "draw the ticket tree", "show the epic on mobile", "add a status graph to

日本語の概要は準備中です。原文の説明を表示しています。

Propose a UI change for human sign-off before building it. The default path deploys a rough live staging preview (with NUXT_PUBLIC_CROUTON_REVIEW=true) so the reviewer pins comments on the real running page. Use --static for the offline HTML/CSS mockup fallback (no deploy available, or speed over fidelity). Invoke for any task that adds/changes a .vue component, a layout, a page, or a theme.

日本語の概要は準備中です。原文の説明を表示しています。

Audit packages for documentation completeness, detect drift between code and docs, and maintain documentation quality across the monorepo. Use when checking package docs, running audits, or reviewing documentation health.

日本語の概要は準備中です。原文の説明を表示しています。

Smart, granular git commits following monorepo conventions. Analyzes changes, filters to session-relevant files, groups by intent, and uses conventional commit format. Use when committing code changes.

日本語の概要は準備中です。原文の説明を表示しています。

The crouton-conformance gate — the SECOND acceptance axis of graduation (the first is the experience side-by-side, /graduate stage C1). Checks a graduated app is a REAL crouton app — CLI-scaffolded, real collections, consumes the correct packages without reinventing them — by composing the existing probes (frontend-review, a11y, e2e-smoke, typecheck) plus crouton-structural checks, then renders a required, signed-off checklist on the graduation PR. Use at /graduate stage C2, or when asked "is this app crouton-native", "run the conformance gate", "check graduation conformance".

日本語の概要は準備中です。原文の説明を表示しています。

The crouton mental model for someone who has never seen this repo — what "layer" means (two different things), the two-halves rule, the three app.config registries, team-scoped multi-tenancy, manifest-driven field types, the numbered invariants with the incidents behind them, and the known-weak points. Use when orienting in this codebase for the first time, when a question is shaped like "how does crouton work / why is it built this way / where does X live", when deciding whether a change belongs in a package or a generated layer, when confused by an extends chain or by croutonApps/croutonCollections/croutonLayoutBlocks, or before proposing any structural change. Trigger phrases: "what is a crouton layer", "package vs generated layer", "how does multi-tenancy work here", "why is there a stub component", "where do field types come from", "what invariants must hold".

日本語の概要は準備中です。原文の説明を表示しています。

The router for how a change legally happens in this repo — which sign-off gate fires for which kind of change, what counts as approval, how the packages-edit guard works, and how work is allowed to reach main. Use when starting ANY change and unsure which rules apply, when an Edit to packages/* gets BLOCKED, when a pipeline issue sits on status:blocked and you need to know what unblocks it, or when asked "do I need sign-off for this", "how do I get this merged", "can I deploy this", "who approves X", "why won't my approval label/reaction do anything".

日本語の概要は準備中です。原文の説明を表示しています。

The case-law chronicle of this repo's major failures — symptom, root cause, evidence, status, and the rule each battle minted — so nobody re-fights a settled battle or re-proposes a rejected fix. Use when a bug/incident looks familiar ("didn't this happen before?"), when you're about to propose an approach that may have been tried and rejected (e.g. shipping a migration from a package), when someone asks "why does rule X exist" / "why is done-by-lgtm a thing" / "what was incident 988", or when a bug-archaeology run hits the shallow-clone wall and needs the pre-June-2026 history. Trigger phrases: "has this broken before", "why can't packages ship migrations", "why did graduation fail", "why is the green-build lesson a rule", "what did we learn from incident NN". (For a live "build is green but the app is broken" symptom, start at crouton-diagnostics-index. NB keep this description free of any space-then-hash sequence — an unquoted YAML scalar treats space-hash as a comment start and silently truncates.)

日本語の概要は準備中です。原文の説明を表示しています。

System reference for the crouton layout engine (@fyit/crouton-layout) — the LayoutTree/layout_configs data model, the croutonLayoutBlocks registry and sizing contract, the viability metric, the deterministic placer (#709), and how layouts persist and seed. Use when reasoning about WHY a layout renders/composes/persists the way it does, when a question is shaped like "what is a LayoutTree / where do layouts live / why did the placer pick this arrangement / why won't this pane shrink / where does minWidth get read", or when a generated POC boots with the wrong (or no) default layout. Trigger phrases: "layout tree", "layout_configs", "viability check", "composeDefaultLayout", "block sizing contract", "why is my layout not viable". For authoring a block component, use the block-authoring skill instead.

日本語の概要は準備中です。原文の説明を表示しています。

How to actually run every kind of test in nuxt-crouton (unit/e2e/typecheck/lint, verified commands) plus the HONEST coverage picture — which packages have real tests vs zero, the dormant skipped/todo cases behind three mocking walls, and what CI truly gates vs what CLAUDE.md claims. Use when asked "how do I run the tests", "is X covered by tests", "what does CI actually check", "why is this suite skipped", "where do I add a test", or before trusting a green build as evidence that something works.

日本語の概要は準備中です。原文の説明を表示しています。

Accessibility review for Vue surfaces — the code-cleaning analog of /code-review and /simplify, pointed at WCAG/ARIA. Reviews just your diff (or a package/file), rates findings by severity, and either comments inline on the PR (--comment) or applies the safe fixes for you (--fix). Steers the depth-aware `a11y` subagent. Use when asked to "check accessibility", "a11y this", "audit ARIA/keyboard", or run /a11y.

日本語の概要は準備中です。原文の説明を表示しています。

The map of nuxt-crouton's GitHub workflows and the deploy-pipeline anatomy — what each workflow gates, how a merge reaches Cloudflare, which secrets/tokens exist where, and how the agent-pipeline workflows chain. Use when a CI check fails and you need to know what it is, when asked "which workflow deploys X / why did staging update / what does this check gate / where is that secret / why didn't my label trigger anything", when wiring a new app into CI, or when debugging a deploy that never fired. Trigger phrases: "what runs on merge to main", "map of the workflows", "WORKER_SECRETS_JSON", "why is the daily sweep not running", "deploy-apps vs deploy-app", "which token does this workflow use".

日本語の概要は準備中です。原文の説明を表示しています。

The catalog of every config-as-data surface in nuxt-crouton — which file controls what, who actually reads it, its drift hazards, and how to change it safely. Use when asking "where is X configured", "who reads this file", "what happens if I edit harness.config.mjs / routing.json / digests.yml / labels.yml / a deploy.config.json / an e2e.manifest.json", when a setting change had no effect (silent no-op), when adding a new config surface, or when a gen-*.mjs --check fails. Trigger phrases: "where do I configure", "what reads this config", "config drift", "the setting didn't do anything", "stale generated doc", "add a new config-as-data surface".

日本語の概要は準備中です。原文の説明を表示しています。

Symptom → root-cause lookup table for nuxt-crouton's recurring failure modes — exact error strings mapped to their known cause, fix, and evidence issue. Use when you hit an error you don't recognize in this repo, before debugging from first principles. Trigger phrases and symptoms: "Cannot resolve entry module .nuxt/hub/db/schema.entry.ts", "Could not load '@fyit/crouton'", "No migrations present", "No schema files found", "$setup.t is not a function", "does not provide an export named 'Map'", "no such table: translations_ui", "collection pages 500 in production", "the build was green but the app is broken", "is this a known issue?".

日本語の概要は準備中です。原文の説明を表示しています。

Which docs in nuxt-crouton to trust and how to resolve contradictions between them. Use when two docs disagree, when a doc's claim doesn't match the code ("the README says packages/nuxt-crouton but that dir doesn't exist"), when deciding whether a guide/writeup is authoritative before following it, or when asked "is this doc current", "which doc wins", "why does CLAUDE.md say X but the code says Y". Also the reference for the supersession-banner practice.

日本語の概要は準備中です。原文の説明を表示しています。

Recreate the nuxt-crouton dev environment from scratch and survive its install/build traps — the pnpm toolchain, the cold-start sequence (why packages must build before apps prepare), the version catalog and workaround pins, the per-app typecheck rule, required env vars, and the known environment traps. Use when setting up a fresh clone/worktree, when "pnpm install" or "nuxt prepare" fails, when an app won't boot after install, when typecheck output looks insane, or when a tool "isn't available". Trigger phrases: "set up the repo from scratch", "why does typecheck show thousands of errors", "playwright install fails", "what pnpm/node version". (Exact error strings — "Could not load '@fyit/crouton'", "Cannot find module '@vue/compiler-sfc'" — route via crouton-diagnostics-index first; it points back here for depth.)

日本語の概要は準備中です。原文の説明を表示しています。

Verified lookup pack for crouton's schema→CRUD generation domain — the fieldsFile JSON format, the complete field-type catalog with aliases, the naming contract, the exact artifact list one generate run produces, the seed/default-layout pipeline, dialect defaults, and rollback semantics. Use when writing or reviewing a collection schema, when asking "what field types exist / what does `crouton config` actually generate / where did this file name come from / why did I get a pg schema", when a generated file's name or path doesn't match the docs, or when removing/regenerating a collection. Trigger phrases — "what field types are supported", "schema JSON format", "what files does crouton generate", "naming convention for collections", "rollback a collection", "dialect default". The generation WORKFLOW (steps + sign-off gates) lives in the `crouton` skill — this is the reference it doesn't carry.

日本語の概要は準備中です。原文の説明を表示しています。

How to MEASURE this repo instead of eyeballing it — interpret the loop-station context-budget trend (history.jsonl, thresholds, the tokenizer-switch gotcha), read a session trace, use the eval ledger (cost-per-success), and pick the right evidence tool (app-shots, smoke-deployed, db-counts, --check modes). Use when asked "is the harness bloating", "how big is CLAUDE.md really", "did this agent run actually succeed", "prove this works", "what evidence do we have", or when tempted to declare a capability (browser, tool) unavailable. For RUNNING the loop-station scripts, defer to the loop-station skill — this skill is interpretation + the wider evidence toolkit.

日本語の概要は準備中です。原文の説明を表示しています。

Author a placeable layout block that looks right at ANY pane size. The one hard rule — size to the PANE with container queries (@container), never the viewport — plus list/form playbooks and the sizing contract (minWidth etc.) the viability metric reads. Use when adding/converting a croutonLayoutBlocks block, or when a block overflows/breaks in a narrow pane.

日本語の概要は準備中です。原文の説明を表示しています。

Day-to-day operation runbook for nuxt-crouton apps — boot any app/POC locally with auth working, seed and inspect the local database (including the miniflare-vs-.data split that makes seeded data "disappear"), turn on the review overlay / eruda, observe deployed staging (tail, smoke, row counts), and capture screenshots. Use when asked to "run the app", "boot velo/fanfare/a poc locally", "why can't I log in locally", "seed the database", "the seed data doesn't show up", "look at the local db", "check what's on staging", "screenshot the app", or "where do the artifacts land".

日本語の概要は準備中です。原文の説明を表示しています。

Emit a blocking question the owner can read in ~10 seconds and answer in one reply — the scannable, recommendation-first handoff every agent posts when it hits a fork it can't own. Leads with the one decision + a recommendation, carries the 🤖 provenance header, doubles as the

日本語の概要は準備中です。原文の説明を表示しています。

Copy one Cloudflare D1 database into another of the same app's environments — refresh STAGING from prod, seed a PREVIEW from staging, or (guarded) push staging up to prod. Full mirror via scripts/db-clone.mjs. Use when asked to "clone a db", "refresh staging from prod", "copy prod data to staging", "seed a preview db", or to reconstitute a wiped staging DB. NOT for per-collection seeding (use the app's seedData/seed endpoints) or for schema migrations (use db-migrations).

日本語の概要は準備中です。原文の説明を表示しています。

The migration step and the package-owned-table exception — NOT the normal collection flow. For app collections use the crouton CLI/skill (schema JSON → `crouton config`); reach here when you then need to generate/apply the Drizzle migration and hit the schema.mjs-only-after-build gotcha ("No schema files found"), or when adding a package-owned infra table (the crouton-flow pattern) that does NOT go through `crouton config`.

日本語の概要は準備中です。原文の説明を表示しています。

When a bug or regression is reported, the FIRST step — before fixing — is to research how and when it was introduced (git archaeology), then record that finding on the tracking issue/PR. Use the moment a bug, error, broken build, or "this used to work" is reported, before writing a fix. Produces a first-bad-commit (or "not a code regression") note you paste onto the issue.

日本語の概要は準備中です。原文の説明を表示しています。