generator-offline-review — for Claude Code generator-offline-review, Mathildanesth, community, for Claude Code, ide skills, generator-ai-review, Skill("generator-offline-review"), — checks, s50-2025, s01-2026

v1.0.0
GitHub

Über diesen Skill

Geeigneter Einsatz: Ideal for AI agents that need generator offline review — no supabase required. Lokalisierte Zusammenfassung: Programme de gestion des plannings équipe d'anesthésie # Generator Offline Review — No Supabase Required Mission : Run the full generator analysis workflow offline using pre-recorded Prisma fixtures. This AI agent skill supports Claude Code, Cursor, and Windsurf

Funktionen

Generator Offline Review — No Supabase Required
Mission : Run the full generator analysis workflow offline using pre-recorded Prisma fixtures.
Works identically to generator-ai-review but without database access.
Invocation : Skill("generator-offline-review") or user says "offline gen" / "génération offline"
Architecture: Record/Replay Prisma Proxy

# Kernthemen

for Claude Code generator-ai-review Skill("generator-offline-review") — checks s50-2025 s01-2026
vperreard vperreard
[1]
[0]
Aktualisiert: 4/24/2026

Skill Overview

Start with fit, limitations, and setup before diving into the repository.

Geeigneter Einsatz: Ideal for AI agents that need generator offline review — no supabase required. Lokalisierte Zusammenfassung: Programme de gestion des plannings équipe d'anesthésie # Generator Offline Review — No Supabase Required Mission : Run the full generator analysis workflow offline using pre-recorded Prisma fixtures. This AI agent skill supports Claude Code, Cursor, and Windsurf

Warum diese Fähigkeit verwenden

Empfehlung: generator-offline-review helps agents generator offline review — no supabase required. Programme de gestion des plannings équipe d'anesthésie # Generator Offline Review — No Supabase Required Mission : Run

Am besten geeignet für

Geeigneter Einsatz: Ideal for AI agents that need generator offline review — no supabase required.

↓ Handlungsfähige Anwendungsfälle for generator-offline-review

Anwendungsfall: Applying Generator Offline Review — No Supabase Required
Anwendungsfall: Applying Mission : Run the full generator analysis workflow offline using pre-recorded Prisma fixtures
Anwendungsfall: Applying Works identically to generator-ai-review but without database access

! Sicherheit & Einschränkungen

  • Einschraenkung: Requires repository-specific context from the skill documentation
  • Einschraenkung: Works best when the underlying tools and dependencies are already configured

About The Source

The section below comes from the upstream repository. Use it as supporting material alongside the fit, use-case, and installation summary on this page.

Labs-Demo

Browser Sandbox Environment

⚡️ Ready to unleash?

Experience this Agent in a zero-setup browser environment powered by WebContainers. No installation required.

Boot Container Sandbox

FAQ und Installationsschritte

These questions and steps mirror the structured data on this page for better search understanding.

? Häufige Fragen

Was ist generator-offline-review?

Geeigneter Einsatz: Ideal for AI agents that need generator offline review — no supabase required. Lokalisierte Zusammenfassung: Programme de gestion des plannings équipe d'anesthésie # Generator Offline Review — No Supabase Required Mission : Run the full generator analysis workflow offline using pre-recorded Prisma fixtures. This AI agent skill supports Claude Code, Cursor, and Windsurf

Wie installiere ich generator-offline-review?

Führen Sie den Befehl aus: npx killer-skills add vperreard/Mathildanesth/generator-offline-review. Er funktioniert mit Cursor, Windsurf, VS Code, Claude Code und mehr als 19 weiteren IDEs.

Wofür kann ich generator-offline-review verwenden?

Wichtige Einsatzbereiche sind: Anwendungsfall: Applying Generator Offline Review — No Supabase Required, Anwendungsfall: Applying Mission : Run the full generator analysis workflow offline using pre-recorded Prisma fixtures, Anwendungsfall: Applying Works identically to generator-ai-review but without database access.

Welche IDEs sind mit generator-offline-review kompatibel?

Dieser Skill ist mit Cursor, Windsurf, VS Code, Trae, Claude Code, OpenClaw, Aider, Codex, OpenCode, Goose, Cline, Roo Code, Kiro, Augment Code, Continue, GitHub Copilot, Sourcegraph Cody, and Amazon Q Developer kompatibel. Nutzen Sie die Killer-Skills CLI für eine einheitliche Installation.

Gibt es Einschränkungen bei generator-offline-review?

Einschraenkung: Requires repository-specific context from the skill documentation. Einschraenkung: Works best when the underlying tools and dependencies are already configured.

↓ So installieren Sie den Skill

  1. 1. Terminal öffnen

    Öffnen Sie Ihr Terminal oder die Kommandozeile im Projektverzeichnis.

  2. 2. Installationsbefehl ausführen

    Führen Sie aus: npx killer-skills add vperreard/Mathildanesth/generator-offline-review. Die CLI erkennt Ihre IDE oder Ihren Agenten automatisch und richtet den Skill ein.

  3. 3. Skill verwenden

    Der Skill ist jetzt aktiv. Ihr KI-Agent kann generator-offline-review sofort im aktuellen Projekt verwenden.

! Source Notes

This page is still useful for installation and source reference. Before using it, compare the fit, limitations, and upstream repository notes above.

Upstream Repository Material

The section below comes from the upstream repository. Use it as supporting material alongside the fit, use-case, and installation summary on this page.

Upstream Source

generator-offline-review

Install generator-offline-review, an AI agent skill for AI agent workflows and automation. Explore features, use cases, limitations, and setup guidance.

SKILL.md
Readonly
Upstream Repository Material
The section below comes from the upstream repository. Use it as supporting material alongside the fit, use-case, and installation summary on this page.
Upstream Source

Generator Offline Review — No Supabase Required

Mission: Run the full generator analysis workflow offline using pre-recorded Prisma fixtures. Works identically to generator-ai-review but without database access.

Invocation: Skill("generator-offline-review") or user says "offline gen" / "génération offline"

Architecture: Record/Replay Prisma Proxy

RECORD (online)                          REPLAY (offline)
┌────────────┐                           ┌────────────┐
│ Real Prisma │──→ Recording Proxy ──→   │ Replay Proxy │
│ (Supabase) │    saves queries to       │ (fixtures)   │
└────────────┘    s50-2025.json          └──────────────┘
                                          ↑ serves from JSON

Files:

  • src/lib/prisma-offline-proxy.ts — Record/Replay proxy
  • src/lib/prisma.ts — checks PRISMA_RECORD_PATH / PRISMA_REPLAY_PATH
  • scripts/offline-fixtures/*.json — fixtures par semaine (s50-2025, s01 à s13-2026, s16-2026)
  • scripts/generate-offline.ts — standalone offline generation (no HTTP, no dev server)
  • scripts/extract-offline-fixtures.ts — records fixtures via direct engine (recommended)
  • scripts/record-offline-fixtures.ts — records fixtures via HTTP generation (legacy)

Semaines disponibles

PresetDatesMoisFixtures
s50-2025 (défaut)2025-12-07 → 2025-12-13Déc 2025offline-fixtures/s50-2025.json ✅
s01-20262025-12-29 → 2026-01-04Janv 2026offline-fixtures/s01-2026.json ✅
s02-20262026-01-05 → 2026-01-11Janv 2026offline-fixtures/s02-2026.json ✅
s03-20262026-01-12 → 2026-01-18Janv 2026offline-fixtures/s03-2026.json ✅
s04-20262026-01-19 → 2026-01-25Janv 2026offline-fixtures/s04-2026.json ✅
s05-20262026-01-26 → 2026-02-01Janv/Fév 2026offline-fixtures/s05-2026.json ✅
s06-20262026-02-02 → 2026-02-08Fév 2026offline-fixtures/s06-2026.json ✅
s07-20262026-02-09 → 2026-02-15Fév 2026offline-fixtures/s07-2026.json ✅
s08-20262026-02-16 → 2026-02-22Fév 2026offline-fixtures/s08-2026.json ✅
s09-20262026-02-23 → 2026-03-01Fév/Mar 2026offline-fixtures/s09-2026.json ✅
s10-20262026-03-02 → 2026-03-08Mar 2026offline-fixtures/s10-2026.json ✅
s11-20262026-03-09 → 2026-03-15Mar 2026offline-fixtures/s11-2026.json ✅
s12-20262026-03-16 → 2026-03-22Mar 2026offline-fixtures/s12-2026.json ✅
s13-20262026-03-23 → 2026-03-29Mar 2026offline-fixtures/s13-2026.json
s16-20262026-04-13 → 2026-04-19Avr 2026offline-fixtures/s16-2026.json ✅

✅ = fixtures enregistrées (S50 + S01-S12 + S16 = 14 semaines). Seule S13 nécessite un enregistrement.

Tous les scripts acceptent --week=<preset> (ex: --week=s10-2026).

Phase 0: CHECK FIXTURES

Before anything, verify fixtures exist for the target week:

bash
1# Vérifier toutes les fixtures disponibles
2ls -la scripts/offline-fixtures/
3
4# Vérifier une semaine spécifique (ex: S10 mars 2026)
5ls -la scripts/offline-fixtures/s10-2026.json

Semaines déjà enregistrées : s50-2025, s01-2026 à s12-2026, s16-2026 (14 semaines). Seule s13-2026 doit encore être enregistrée si nécessaire.

If the file does not exist or needs refreshing:

Record fixtures (requires Supabase access):

Method A: Standalone script (recommended, no server restart)

bash
1# Enregistrer une semaine spécifique (ex: S10 mars 2026)
2npx tsx --env-file .env.local scripts/extract-offline-fixtures.ts --week=s10-2026
3
4# Enregistrer plusieurs semaines en séquence
5for week in s02-2026 s06-2026 s10-2026; do
6  npx tsx --env-file .env.local scripts/extract-offline-fixtures.ts --week=$week
7done

This runs the engine directly with the recording proxy. No dev server needed.

Method B: Via dev server (if you need more control)

  1. Stop any running dev server
  2. Start with recording: 
    bash
    1PRISMA_RECORD_PATH=scripts/offline-fixtures/s10-2026.json npm run dev
  3. Run a generation: 
    bash
    1npx tsx --env-file .env.local scripts/generate-via-api.ts --week=s10-2026
  4. Stop the dev server (Ctrl+C)

Verify the fixtures file was created and has reasonable size (>1MB = ~170 queries).

Phase 1: OFFLINE GENERATE

Two options:

bash
1# S50 (défaut)
2npx tsx --env-file .env.local scripts/generate-offline.ts
3npx tsx --env-file .env.local scripts/generate-offline.ts --hybrid=false
4npx tsx --env-file .env.local scripts/generate-offline.ts --bestOfN=3
5
6# S16
7npx tsx --env-file .env.local scripts/generate-offline.ts --week=s16-2026
8npx tsx --env-file .env.local scripts/generate-offline.ts --week=s16-2026 --bestOfN=3

This imports the engine directly with PRISMA_REPLAY_PATH set. No Next.js, no HTTP, no dev server required.

Option B: Via dev server (if you need the UI or HTTP endpoints)

bash
1# S50
2PRISMA_REPLAY_PATH=scripts/offline-fixtures/s50-2025.json npm run dev
3npx tsx --env-file .env.local scripts/generate-via-api.ts
4
5# S16
6PRISMA_REPLAY_PATH=scripts/offline-fixtures/s16-2026.json npm run dev
7npx tsx --env-file .env.local scripts/generate-via-api.ts --week=s16-2026

After generation:

  1. Read scripts/debug-exports/runs/<logId>/validation.json (archivé, figé, recommandé depuis 24/04/2026) OU scripts/debug-exports/latest-validation.json (risque d'écrasement au prochain run)
  2. Generate the visual grid: 
    bash
    1npm run debug:grid
2# équivalent : npx tsx --env-file .env.local scripts/view-generation-grid.ts --no-open

🆕 Archivage automatique par logId (24/04/2026)

Chaque génération crée automatiquement scripts/debug-exports/runs/<logId>/ contenant :

  • manifest.json — summary_for_ai (5 lignes), overallScore, topViolations, stats
  • validation.json, assignments.json, generation-trace.json, r1-analysis.json, surgeon-grid.json, staffing-plan.json
  • FIFO 20 runs max (anciens supprimés automatiquement)

Pour comparer 2 runs offline (avant fix vs après fix) :

typescript
1import { DebugAutoExporter } from '@/modules/unified-generator/services/DebugAutoExporter';
2const runs = DebugAutoExporter.listRuns();
3const before = DebugAutoExporter.getRunManifest(runs[1].logId);
4const after = DebugAutoExporter.getRunManifest(runs[0].logId);
5console.log(before.summary_for_ai, '\n→\n', after.summary_for_ai);

Output: Display quickSummary + structuredSummary + manifest.summary_for_ai du run archivé.

Phase 2+3: ANALYSIS (Claude + Codex)

Identical to generator-ai-review Phase 2+3.

Use the same analysis workflow:

  • Read scripts/debug-exports/latest-validation.json
  • Perform 2A (integrity check), 2B (causal analysis), 2C (prioritization), 2D (code reading)
  • Launch Codex analysis in parallel (if codex CLI available)

Note: In offline mode, Codex may or may not be available. If no internet → skip Codex, do Claude-only analysis. If internet available (VPN, hotspot) → use Codex normally.

Source File Mapping (same as generator-ai-review)

RulesSource File
R1,R4,R5,R6,R7,R10,R19,R22,R23,R27,R36phases/IADEDistributionPhase.ts
R9,R11,R12,R13,R18,R48phases/MARBlocPhase.ts
R8,R14,R15,R28,R30phases/SupervisionPhase.ts
R12,R26,R41phases/GlobalSwapOptimizerPhase.ts
R12,R33,R41phases/CrossDayOptimizerPhase.ts
R21,R32phases/OffDistributionPhase.ts
R16,R17,R29,R31,R44,R46phases/GardePhase.ts
R42phases/AstreintePhase.ts
R20,R25,R43,R47phases/ConsultationPhase.ts

Phase 4: CROSS-VALIDATION

Same as generator-ai-review Phase 4:

  • Triple validation (Report + Claude + Codex if available)
  • Merge findings
  • Filter justified/period-limited violations

Phase 5: JOINT FIX PLAN

Same format as generator-ai-review Phase 5. Present to user for approval.

Phase 6: EXECUTE FIXES + RE-GENERATE OFFLINE

For each approved fix:

  1. Read the target file
  2. Apply the fix via Edit tool
  3. Re-generate offline: 
    bash
    1npx tsx --env-file .env.local scripts/generate-offline.ts
  4. Compare before/after scores
  5. If regression (>2pts): revert

After all fixes:

  • Run 3x: npx tsx --env-file .env.local scripts/generate-offline.ts --bestOfN=1
  • Present before/after table
  • Update memory files

Guardrails

Same as generator-ai-review:

  • Never modify the Prisma schema
  • Never change validation rules to mask violations
  • Never do mutating requests against localhost:3000

Offline-specific:

  • Always verify fixtures exist before generating
  • Remember: Fixtures are a snapshot — if DB data changes, re-record with --week=<preset>
  • N=3 validation still works: Each run is stochastic (shuffleRooms), fixtures just provide the input data
  • CasePackExporter may partially fail in direct engine mode (some export queries not recorded). The assignments and core validation still work.

Quick Reference

🆕 Scripts npm (depuis 24/04/2026 — préférés pour la simplicité)

bash
1npm run debug:offline              # Génération offline S50 (défaut)
2npm run debug:offline -- --week=s16-2026    # Semaine spécifique
3npm run debug:grid                 # Génère generation-grid.html
4npm run debug:phases               # Génère phase-visualizer.html
5npm run debug:latest               # Dump des latest-*.json dans le terminal
6npm run debug:trace -- "NOM" "2026-04-13" "MATIN" "Salle 2"   # Trace user/slot

Commandes complètes (legacy — équivalentes)

bash
1# Record fixtures (online, requires Supabase) — à faire UNE FOIS par semaine
2npx tsx --env-file .env.local scripts/extract-offline-fixtures.ts                   # S50 2025 (défaut)
3npx tsx --env-file .env.local scripts/extract-offline-fixtures.ts --week=s02-2026   # Janv 2026
4npx tsx --env-file .env.local scripts/extract-offline-fixtures.ts --week=s06-2026   # Fév 2026
5npx tsx --env-file .env.local scripts/extract-offline-fixtures.ts --week=s10-2026   # Mars 2026
6npx tsx --env-file .env.local scripts/extract-offline-fixtures.ts --week=s16-2026   # Avr 2026
7
8# Generate offline (no server needed) — toutes semaines
9npx tsx --env-file .env.local scripts/generate-offline.ts                   # S50 (défaut)
10npx tsx --env-file .env.local scripts/generate-offline.ts --week=s02-2026   # Janv
11npx tsx --env-file .env.local scripts/generate-offline.ts --week=s06-2026   # Fév
12npx tsx --env-file .env.local scripts/generate-offline.ts --week=s10-2026   # Mars
13npx tsx --env-file .env.local scripts/generate-offline.ts --week=s16-2026   # Avr
14
15# Generate offline via dev server
16PRISMA_REPLAY_PATH=scripts/offline-fixtures/s50-2025.json npm run dev
17npx tsx --env-file .env.local scripts/generate-via-api.ts
18# Ou pour une semaine spécifique:
19PRISMA_REPLAY_PATH=scripts/offline-fixtures/s10-2026.json npm run dev
20npx tsx --env-file .env.local scripts/generate-via-api.ts --week=s10-2026
21
22# View grid
23npx tsx scripts/view-generation-grid.ts --no-open
24
25# 🆕 Pack IA unifié (nécessite dev server actif, online OU replay mode)
26curl "http://localhost:3000/api/debug-generator/ai-pack?logId=<logId>" > /tmp/ai-pack.json

Verwandte Fähigkeiten

Looking for an alternative to generator-offline-review or another community skill for your workflow? Explore these related open-source skills.

Alle anzeigen

openclaw-release-maintainer

Logo of openclaw
openclaw

Lokalisierte Zusammenfassung: 🦞 # OpenClaw Release Maintainer Use this skill for release and publish-time workflow. It covers ai, assistant, crustacean workflows. This AI agent skill supports Claude Code, Cursor, and Windsurf workflows.

★ 333.8k
⑂ 0
Künstliche Intelligenz

widget-generator

Logo of f
f

Lokalisierte Zusammenfassung: Generate customizable widget plugins for the prompts.chat feed system # Widget Generator Skill This skill guides creation of widget plugins for prompts.chat . It covers ai, artificial-intelligence, awesome-list workflows. This AI agent skill supports Claude Code

★ 149.6k
⑂ 0
Künstliche Intelligenz

flags

Logo of vercel
vercel

Lokalisierte Zusammenfassung: The React Framework # Feature Flags Use this skill when adding or changing framework feature flags in Next.js internals. It covers blog, browser, compiler workflows. This AI agent skill supports Claude Code, Cursor, and Windsurf workflows.

★ 138.4k
⑂ 0
Browser

pr-review

Logo of pytorch
pytorch

Lokalisierte Zusammenfassung: Usage Modes No Argument If the user invokes /pr-review with no arguments, do not perform a review . It covers autograd, deep-learning, gpu workflows. This AI agent skill supports Claude Code, Cursor, and Windsurf workflows.

★ 98.6k
⑂ 0
Entwickler