@milenyumai/film-kit 1.0.3 → 1.0.4

package/README.md

@@ -1,89 +1,95 @@
-# ShotForge Agent Configuration
+# Film-Kit: Hollywood-Standard Cinematic Prompt Engineering
 
-ShotForge artık bir **agent configuration npm paketi**dir.
-Amaç: Cursor, Claude Code, VS Code Copilot ve Google Antigravity için proje içinde doğru kural/skill dosyalarını otomatik oluşturmak.
+**Film-Kit**, Google Veo 3.1 ve Flow için optimize edilmiş, profesyonel "Prompt Engineering" ajanlarını (Claude Code, Cursor) projenize kuran bir araç setidir.
 
-Bu paket **shot üreten bir CLI değildir**.
-Editördeki AI ajanlarına `/generate` davranışını öğreten yapı dosyalarını yazar.
+İki versiyonu vardır:
+1. **Film-Kit (Core):** Tek ajanlı, manuel kontrollü üretim.
+2. **Film-Kit Multi (Pro):** Çok ajanlı, paralel üretim ve uzman denetimi.
 
-## Ne Oluşturur?
+---
 
-### Cursor
-- `./.cursorrules` (legacy)
-- `./.cursor/rules/global.mdc`
-- (ortak sözleşme) `./AGENTS.md`
-
-### Claude Code
-- `./CLAUDE.md`
-- `./.claude/CLAUDE.md`
-- `./.claude/rules/generate-flow.md`
-- `./.claude/rules/output-contract.md`
-- `./.claude/settings.json`
-- `./.claude/settings.local.json`
-
-### VS Code Copilot
-- `./.github/copilot-instructions.md`
-- `./.github/instructions/shotforge.instructions.md`
-
-### Google Antigravity
-- `./.agent/skills/shotforge-generate/SKILL.md`
-
-## Kurulum
+## 📦 Paket 1: Film-Kit (Core)
+**Kullanım:** Bireysel projeler, öğrenme, <10 shot senaryolar.
 
+### Kurulum (Tek Komut)
+Boş bir klasörde terminali açın:
 ```bash
-npm i -D @your-scope/shotforge-agent-config
+npx @milenyumai/film-kit init
 ```
 
-Kurulumdan sonra paket `postinstall` ile proje köküne dosyaları yazmaya çalışır.
-
-## Opsiyonel Proje Konfigürasyonu
-
-Proje köküne `shotforge-agent.config.json` koyarsan varsayılanları özelleştirebilirsin:
-
-```json
-{
-  "outputDir": "./shots",
-  "scenarioHint": "scenario.md",
-  "platforms": ["cursor", "claude", "copilot", "antigravity"],
-  "overwrite": false,
-  "includeAgentsMd": true
-}
+### Kullanım
+Claude Code içinde:
+```text
+/generate
 ```
+Tek bir ajan ("Prompt Engineer") senaryoyu okur, analiz eder ve shotları sırayla üretir.
 
-## Programatik Kullanım (CLI yok)
+**Özellikler:**
+- Veo 3.1 Prompt Flow (Subject > Action > Camera > Light)
+- Profesyonel Audio Blocks (Dialogue, SFX, Mix)
+- Zorunlu Avoid Line
+- Auto-Safety & Auto-Anonymous
 
-İstersen Node içinden direkt çağırabilirsin:
+---
 
-```ts
-import { configureAgents } from "@your-scope/shotforge-agent-config";
+## 🚀 Paket 2: Film-Kit Multi (Pro)
+**Kullanım:** Profesyonel stüdyo işleri, uzun metraj, 10+ shot senaryolar.
 
-await configureAgents({
-  rootDir: process.cwd(),
-  outputDir: "./shots",
-  scenarioHint: "scenario.md",
-  platforms: ["cursor", "claude", "copilot", "antigravity"],
-  overwrite: false
-});
+### Kurulum (Tek Komut)
+Boş bir klasörde terminali açın:
+```bash
+npx @milenyumai/film-kit-multi init
 ```
 
-## Editör Davranış Sözleşmesi
-
-Yazılan dosyalar editör AI'ına şu kontratı verir:
-- Kullanıcı `/generate` dediğinde seçili/aktif senaryo dosyasını oku.
-- `./shots/bible.md` ve `./shots/S01.md ...` formatında markdown çıktısı üret.
-- Shot dosyalarında prompt + cinematography + continuity bilgisi olsun.
-
-## Global Skill Notu (Antigravity)
-
-Bu paket varsayılan olarak sadece workspace içine yazar.
-Global skill yolu bilgi amaçlı:
-- `~/.gemini/antigravity/skills/<skill-adı>/SKILL.md`
-
-## Geliştirme
-
-```bash
-npm install
-npm run typecheck
-npm test
-npm run build
+### Kullanım
+Claude Code içinde:
+```text
+/generate-multi
 ```
+**Lead Director** ajanı ekibi kurar ve yönetir.
+
+**Süreç:**
+1. **Planning:** Lead Director senaryoyu analiz eder, `team-plan.json` oluşturur.
+2. **Casting:** İhtiyaca göre 5'e kadar **Shot Generator** ajanı spawn eder.
+3. **Execution:** Ajanlar shotları (Min 60 kelime + coverage) paralel üretir.
+4. **Validation:** 3 Uzman Ajan devreye girer:
+   - `continuity-editor`: Zincirleme (chain) hatalarını yakalar.
+   - `safety-auditor`: Güvenlik ve anonimlik ihlallerini tarar.
+   - `delivery-editor`: Dosya bütünlüğünü ve formatı denetler.
+5. **Recovery:** Hata varsa `/recover` workflow'u otomatik düzeltir.
+
+---
+
+## ⚡ Komut Referansı (Slash Commands)
+
+| Komut | Paket | İşlev |
+|-------|-------|-------|
+| `/generate` | Core | Tek ajanla sıralı üretim başlatır. |
+| `/generate-multi` | Multi | Multi-ajan takımıyla paralel üretim başlatır. |
+| `/chain` | Her İkisi | Son üretilen shot'ın bitiş karesinden devam eder. |
+| `/finish` | Her İkisi | Projeyi sonlandırır ve `FINAL-SUMMARY.md` yazar. |
+| `/recover` | Multi | Başarısız batch'leri veya kalite kontrol takılanları onarır. |
+| `/safety-check` | Her İkisi | Manuel güvenlik taraması yapar. |
+
+---
+
+## 🆚 Karşılaştırma
+
+| Özellik | Film-Kit (Core) | Film-Kit Multi (Pro) |
+|---------|-----------------|----------------------|
+| **Mimari** | Tek Ajan | Lead + Teammates + Specialists |
+| **Hız (20 Shot)** | ~60 dk | ~20 dk (Paralel) |
+| **Tutarlılık** | Manuel Kontrol | `continuity-editor` Denetimi |
+| **Güvenlik** | Prompt İçi Filtre | `safety-auditor` Taraması |
+| **Planlama** | Yok | `team-plan.json` Kontratlı |
+| **Recovery** | Manuel | Otomatik `/recover` |
+
+---
+
+## 🛠 Gereksinimler
+- **Node.js:** v18+
+- **Claude Code:** `CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1` (Sadece Multi için)
+- **Editör:** VS Code, Cursor veya Terminal
+
+## 📄 Lisans
+MIT

package/build/cli.js

@@ -2,10 +2,14 @@
 import { configureAgents } from "./lib/configure.js";
 const args = process.argv.slice(2);
 const command = args[0];
+const overwrite = args.includes("--overwrite") || args.includes("-f");
 if (command === "init") {
     console.log("🎬 Film-Kit: Configuring AI agents...\n");
     try {
-        const result = await configureAgents({ rootDir: process.cwd() });
+        const result = await configureAgents({
+            rootDir: process.cwd(),
+            overwrite
+        });
         if (result.written.length > 0) {
             console.log(`✅ Files written: ${result.written.length}`);
             result.written.forEach(f => console.log(`   ${f}`));
@@ -26,7 +30,8 @@ else {
 🎬 @milenyumai/film-kit — Hollywood-grade prompt engineering for Veo 3.1
 
 Usage:
-  npx @milenyumai/film-kit init    Configure AI agents in current directory
+  npx @milenyumai/film-kit init                 Configure AI agents in current directory
+  npx @milenyumai/film-kit init --overwrite     Force update existing .agent content files
 
 Or install as a dependency (auto-configures on install):
   npm install --save-dev @milenyumai/film-kit

package/build/lib/templates.js

@@ -57,12 +57,16 @@ All rules, skills, and workflows are located under \`.agent/\`.
 | \`/generate\` | \`.agent/workflows/generate.md\` |
 | \`/chain\` | \`.agent/workflows/chain.md\` |
 | \`/safety-check\` | \`.agent/workflows/safety-check.md\` |
+| \`/recover\` | \`.agent/workflows/recover.md\` |
 | \`/finish\` | \`.agent/workflows/finish.md\` |
 
 ## Goal
 When the user asks \`/generate\`, convert the scenario into:
 - \`${config.outputDir}/project-info.md\` — Characters, settings, arc mapping
+- \`${config.outputDir}/shot-plan.json\` — Single-agent plan + policy contract
 - \`${config.outputDir}/shots/SHOT01.md, SHOT02.md, ...\` — Production shot files (with coverage included)
+- \`${config.outputDir}/reports/SAFETY-REPORT.md\` — Safety gate result
+- \`${config.outputDir}/reports/DELIVERY-REPORT.md\` — Delivery gate result
 - \`${config.outputDir}/_index.md\` — Shot list with chain & status tracking
 
 ## Input
@@ -81,6 +85,7 @@ Each \`SHOTNN.md\` is a **single file** containing ALL shot details:
 
 ## Critical Rules
 - **AUTO-ANONYMOUS:** Replace ALL real person names with physical descriptions automatically
+- **Name Policy:** Visual prompts must stay anonymous. Dialogue naming follows \`shot-plan.json\` policy.
 - **AUTO-SAFETY:** Proactively reframe content that may trigger safety filters
 - **Frame Chaining:** Last frame of SHOT[N] = First frame of SHOT[N+1]
 - **Coverage Mandatory:** Every main shot includes 2-3 coverage sub-shots in same file
@@ -107,20 +112,23 @@ Before generating ANY prompts, read the appropriate skill files from .agent/skil
 | Prompt Structure | .agent/skills/prompt-structure/SKILL.md | ALWAYS |
 
 ## UNIVERSAL RULES
-1. AUTO-ANONYMOUS: Replace ALL real person names with physical descriptions. NEVER use celebrity names.
-2. AUTO-SAFETY: Proactively reframe content that may trigger safety filters.
-3. Prompts ALWAYS in English. Headers in Turkish allowed. Dialogue preserved in original language.
-4. SLOW BURN: 8s default duration. Split actions into multiple shots. "Ilmek ilmek islemek."
-5. Music: NONE by default. User must explicitly request.
-6. EVERY prompt must have an Avoid line. No exceptions.
-7. Coverage shots mandatory (2-3 per main shot, min 60 words each, included in same file).
-8. Frame chaining: Last frame of SHOT[N] = First frame of SHOT[N+1].
-9. ONE FILE PER SHOT: Each SHOTNN.md contains main shot + all coverage shots.
+1. AUTO-ANONYMOUS: Replace ALL real person names in visual prompts with physical descriptions. NEVER use celebrity names.
+2. Dialogue Name Policy: Follow ${config.outputDir}/shot-plan.json -> dialogue_name_policy.
+3. AUTO-SAFETY: Proactively reframe content that may trigger safety filters.
+4. Prompts ALWAYS in English. Headers in Turkish allowed. Dialogue preserved in original language.
+5. SLOW BURN: 8s default duration. Split actions into multiple shots. "Ilmek ilmek islemek."
+6. Music: NONE by default. User must explicitly request.
+7. EVERY prompt must have an Avoid line. No exceptions.
+8. Coverage shots mandatory (2-3 per main shot, min 60 words each, included in same file).
+9. Frame chaining: Last frame of SHOT[N] = First frame of SHOT[N+1].
+10. ILK/İLK FRAME section must contain a code block even for chained shots.
+11. ONE FILE PER SHOT: Each SHOTNN.md contains main shot + all coverage shots.
 
 ## WORKFLOWS
 - /generate → Read .agent/workflows/generate.md
 - /chain → Read .agent/workflows/chain.md
 - /safety-check → Read .agent/workflows/safety-check.md
+- /recover → Read .agent/workflows/recover.md
 - /finish → Read .agent/workflows/finish.md
 
 ## OUTPUT
@@ -157,16 +165,19 @@ All skills at: \`.agent/skills/[name]/SKILL.md\`
 | /generate | Generate shots → Read \`.agent/workflows/generate.md\` |
 | /chain | Continue from last → Read \`.agent/workflows/chain.md\` |
 | /safety-check | Validate prompts → Read \`.agent/workflows/safety-check.md\` |
+| /recover | Recover failed gates → Read \`.agent/workflows/recover.md\` |
 | /finish | Complete project → Read \`.agent/workflows/finish.md\` |
 
 ## CRITICAL RULES
 - AUTO-ANONYMOUS: Replace ALL real names with physical descriptions
+- Dialogue naming follows \`${config.outputDir}/shot-plan.json\` policy
 - AUTO-SAFETY: Proactively reframe sensitive content
 - Frame chaining: Last frame SHOT[N] = First frame SHOT[N+1]
 - Coverage: 2-3 sub-shots per main shot (min 60 words each, in same file)
 - Avoid line: MANDATORY on every prompt
 - Music: NONE by default
 - Duration: 8s default, slow burn pacing
+- ILK/İLK FRAME section always contains a code block
 - ONE FILE PER SHOT: SHOTNN.md contains everything
 
 ## Output
@@ -197,14 +208,16 @@ Before generating ANY prompts, read skills from \`.agent/skills/\`:
 - \`prompt-structure/SKILL.md\` — ALWAYS (P5)
 
 ## Universal Rules (Always Active)
-1. **AUTO-ANONYMOUS:** Replace ALL real person names with physical descriptions. NEVER use celebrity names.
-2. **AUTO-SAFETY:** Proactively reframe content that may trigger safety filters.
-3. **Language:** Prompts ALWAYS in English. Dialogue preserved in original.
-4. **Pacing:** 8s default, slow burn. "Ilmek ilmek islemek."
-5. **Music:** NONE by default.
-6. **Avoid line:** MANDATORY on every prompt (image, video, coverage).
-7. **Coverage:** 2-3 sub-shots within same SHOTNN.md file, min 60 words each.
-8. **ONE FILE PER SHOT:** No separate coverage files.
+1. **AUTO-ANONYMOUS:** Replace ALL real person names in visual prompts with physical descriptions.
+2. **Dialogue Name Policy:** Follow \`${config.outputDir}/shot-plan.json\` policy.
+3. **AUTO-SAFETY:** Proactively reframe content that may trigger safety filters.
+4. **Language:** Prompts ALWAYS in English. Dialogue preserved in original.
+5. **Pacing:** 8s default, slow burn. "Ilmek ilmek islemek."
+6. **Music:** NONE by default.
+7. **Avoid line:** MANDATORY on every prompt (image, video, coverage).
+8. **Coverage:** 2-3 sub-shots within same SHOTNN.md file, min 60 words each.
+9. **ILK/İLK FRAME:** Always include a fenced code block, even when chained.
+10. **ONE FILE PER SHOT:** No separate coverage files.
 
 ## Workflows
 | Command | Workflow |
@@ -212,6 +225,7 @@ Before generating ANY prompts, read skills from \`.agent/skills/\`:
 | \`/generate\` | \`.agent/workflows/generate.md\` |
 | \`/chain\` | \`.agent/workflows/chain.md\` |
 | \`/safety-check\` | \`.agent/workflows/safety-check.md\` |
+| \`/recover\` | \`.agent/workflows/recover.md\` |
 | \`/finish\` | \`.agent/workflows/finish.md\` |
 
 ## I/O
@@ -219,6 +233,8 @@ Before generating ANY prompts, read skills from \`.agent/skills/\`:
 - Output: \`${config.outputDir}/shots/SHOTNN.md\` per shot (single file with coverage)
 - Index: \`${config.outputDir}/_index.md\`
 - Project info: \`${config.outputDir}/project-info.md\`
+- Plan: \`${config.outputDir}/shot-plan.json\`
+- Reports: \`${config.outputDir}/reports/SAFETY-REPORT.md\`, \`${config.outputDir}/reports/DELIVERY-REPORT.md\`
 
 Follow rules under \`.claude/rules/\`.
 `;
@@ -234,7 +250,9 @@ All production rules are in \`.agent/MASTER.md\`.
 2. Analyze scenario, auto-anonymize, apply safety
 3. Generate shot files at \`${config.outputDir}/shots/SHOTNN.md\`
 4. Each SHOTNN.md: İLK FRAME + SON FRAME + VİDEO + 2-3 Coverage (ALL IN ONE FILE)
-5. Maintain frame chaining continuity
+5. Keep İLK FRAME as a fenced code block even when chained
+6. Maintain visual anonymity; apply dialogue name policy from \`${config.outputDir}/shot-plan.json\`
+7. Generate/refresh reports in \`${config.outputDir}/reports/\` before finish
 
 ## Priority Hierarchy
 | Priority | Rule |
@@ -256,10 +274,12 @@ Read \`.agent/workflows/generate.md\` for the complete generation workflow.
 1. **Scenario Analysis:**
    - Read scenario, auto-detect genre
    - Identify characters and locations
-   - AUTO-ANONYMOUS: Replace all real names with physical descriptions
+   - AUTO-ANONYMOUS: Replace all real names in visual prompts with physical descriptions
+   - Dialogue names follow \`${config.outputDir}/shot-plan.json\` policy
    - AUTO-SAFETY: Proactively reframe sensitive content
    - Determine shot count based on action beats
    - Create \`${config.outputDir}/project-info.md\`
+   - Create \`${config.outputDir}/shot-plan.json\`
 
 2. **Batch Strategy:**
    - 1-10 shots → Generate all at once
@@ -269,11 +289,18 @@ Read \`.agent/workflows/generate.md\` for the complete generation workflow.
 3. **Per Shot (SINGLE FILE: SHOTNN.md):**
    - Analyze scene type (Dialogue / Action / Emotional / Establishing)
    - Generate main shot (İLK FRAME + SON FRAME + VİDEO)
+   - Keep İLK FRAME as fenced code block even when chained
    - Generate 2-3 coverage shots (in same file)
    - Write to \`${config.outputDir}/shots/SHOT[NN].md\`
    - Update \`${config.outputDir}/_index.md\`
 
-4. **Coverage Selection Matrix:**
+4. **Validation Gates:**
+   - Run /safety-check
+   - Write \`${config.outputDir}/reports/SAFETY-REPORT.md\`
+   - Write \`${config.outputDir}/reports/DELIVERY-REPORT.md\`
+   - If any gate fails, run \`.agent/workflows/recover.md\`
+
+5. **Coverage Selection Matrix:**
    - Dialogue → Reaction (A) + OTS (B)
    - Action → Insert (A) + ECU (B)
    - Emotional → Reaction (A) + ECU (B)
@@ -287,8 +314,11 @@ function buildClaudeRuleOutputContract(config) {
 
 ## Required Files
 - \`${config.outputDir}/project-info.md\` — Characters, settings, emotional arc mapping, tension levels
+- \`${config.outputDir}/shot-plan.json\` — Name policy, shot plan, and validation contract
 - \`${config.outputDir}/_index.md\` — Shot tracking with chain & status
 - \`${config.outputDir}/shots/SHOT01.md ... SHOTNN.md\` — Individual shot files (one file per shot)
+- \`${config.outputDir}/reports/SAFETY-REPORT.md\` — Safety gate report
+- \`${config.outputDir}/reports/DELIVERY-REPORT.md\` — Delivery gate report
 
 ## Prompt Flow Order (Veo Optimization — MANDATORY)
 
@@ -320,6 +350,9 @@ FIRST SHOT / CHAINED from SHOT[prev]_END / CHAIN BREAK - Reason
 ### İLK FRAME (SHOTNN_START)
 [If chained: "→ Use SHOT[prev]_END as first frame"]
 
+> NOTE: Even when chained, this section MUST contain a fenced code block.
+> If chained, include: "Use SHOT[prev]_END as exact first frame."
+
 \\\`\\\`\\\`
 [Image prompt — min 60 words, following prompt flow order]
 Avoid: blurry, low-res, noise, distorted faces, bad anatomy, extra limbs/fingers,
@@ -429,6 +462,10 @@ Total: 1 main + [N] coverage = [N+1] production shots
 - Main shots chain: SON FRAME of SHOT[N] = İLK FRAME of SHOT[N+1]
 - Coverage exists in parallel, not in sequence
 
+### Name Rule
+- Visual prompts and non-dialogue fields: no real names
+- Dialogue transcript naming follows \`shot-plan.json.dialogue_name_policy\`
+
 ### 30° Rule
 Coverage camera angles MUST differ from main shot by at least 30°.
 Insert shots are exempt (different subject entirely).
@@ -490,9 +527,12 @@ Before generating ANY prompts, read skills from \`.agent/skills/\`:
 3. Generate shot files: \`${config.outputDir}/shots/SHOT01.md, SHOT02.md, ...\`
 4. Create index: \`${config.outputDir}/_index.md\`
 5. Create project info: \`${config.outputDir}/project-info.md\`
+6. Create plan: \`${config.outputDir}/shot-plan.json\`
+7. Write reports: \`${config.outputDir}/reports/SAFETY-REPORT.md\`, \`${config.outputDir}/reports/DELIVERY-REPORT.md\`
 
 ### Critical Rules
 - **AUTO-ANONYMOUS:** Replace ALL real names with physical descriptions
+- **Name Policy:** Dialogue naming follows \`${config.outputDir}/shot-plan.json\` policy
 - **AUTO-SAFETY:** Proactively reframe sensitive content
 - **Frame Chaining:** Last frame of SHOT[N] = First frame of SHOT[N+1]
 - **Coverage:** 2-3 sub-shots per main shot (in same file, min 60 words each)
@@ -500,11 +540,13 @@ Before generating ANY prompts, read skills from \`.agent/skills/\`:
 - **Music:** NONE by default
 - **Duration:** 8s default, slow burn pacing
 - **Language:** Prompts in English, dialogue preserved
+- **ILK/İLK FRAME:** keep fenced code block even when chained
 - **ONE FILE PER SHOT:** SHOTNN.md contains main shot + all coverage
 
 ### Other Workflows
 - \`/chain\` → \`.agent/workflows/chain.md\`
 - \`/safety-check\` → \`.agent/workflows/safety-check.md\`
+- \`/recover\` → \`.agent/workflows/recover.md\`
 - \`/finish\` → \`.agent/workflows/finish.md\`
 
 ### Input
@@ -521,9 +563,10 @@ When request is /generate, follow the Film-Kit Hollywood production system:
 1. Read \`.agent/workflows/generate.md\` for full procedure
 2. Load required skills from \`.agent/skills/\`
 3. Transform scenario into production shot package at \`${config.outputDir}\`
-4. Generate: project-info.md, _index.md, shots/SHOT01.md..SHOTNN.md
+4. Generate: project-info.md, shot-plan.json, _index.md, shots/SHOT01.md..SHOTNN.md
 5. Each SHOTNN.md: İLK FRAME + SON FRAME + VİDEO + 2-3 Coverage (ALL IN ONE FILE)
-6. Enforce: auto-anonymous, auto-safety, frame chaining, avoid lines
+6. Enforce: auto-anonymous, dialogue name policy, auto-safety, frame chaining, avoid lines
+7. Write reports to \`${config.outputDir}/reports/\` before /finish
 `;
 }
 /* ---------- ANTIGRAVITY ---------- */
@@ -555,6 +598,7 @@ Before generating ANY prompts, read these skills:
 | /generate | \`.agent/workflows/generate.md\` — Generate shots from scenario |
 | /chain | \`.agent/workflows/chain.md\` — Continue from last shot |
 | /safety-check | \`.agent/workflows/safety-check.md\` — Validate before delivery |
+| /recover | \`.agent/workflows/recover.md\` — Recover failed gates |
 | /finish | \`.agent/workflows/finish.md\` — Complete project, create summary |
 
 ## Input
@@ -565,17 +609,21 @@ Before generating ANY prompts, read these skills:
 - Per-shot files: \`${config.outputDir}/shots/SHOTNN.md\` (single file per shot with coverage)
 - Shot index: \`${config.outputDir}/_index.md\`
 - Project info: \`${config.outputDir}/project-info.md\`
+- Plan: \`${config.outputDir}/shot-plan.json\`
+- Reports: \`${config.outputDir}/reports/SAFETY-REPORT.md\`, \`${config.outputDir}/reports/DELIVERY-REPORT.md\`
 
 ## Critical Rules
 1. **AUTO-ANONYMOUS:** Replace ALL real person names with physical descriptions
-2. **AUTO-SAFETY:** Proactively reframe sensitive content
-3. **Frame Chaining:** Last frame of SHOT[N] becomes first frame of SHOT[N+1]
-4. **Coverage Mandatory:** 2-3 sub-shots per main shot (in same file, min 60 words each)
-5. **Avoid Line:** MANDATORY on every prompt (image + video + coverage)
-6. **Music: NONE** by default
-7. **Ultra Realism** default visual mode
-8. **8s duration** default, slow burn pacing
-9. **ONE FILE PER SHOT:** SHOTNN.md contains main shot + all coverage
+2. **Name Policy:** Dialogue naming follows \`${config.outputDir}/shot-plan.json\` policy
+3. **AUTO-SAFETY:** Proactively reframe sensitive content
+4. **Frame Chaining:** Last frame of SHOT[N] becomes first frame of SHOT[N+1]
+5. **Coverage Mandatory:** 2-3 sub-shots per main shot (in same file, min 60 words each)
+6. **Avoid Line:** MANDATORY on every prompt (image + video + coverage)
+7. **Music: NONE** by default
+8. **Ultra Realism** default visual mode
+9. **8s duration** default, slow burn pacing
+10. **ILK/İLK FRAME:** always keep fenced code block
+11. **ONE FILE PER SHOT:** SHOTNN.md contains main shot + all coverage
 
 ## Quality Standard
 HOLLYWOOD PRODUCTION GRADE — Every frame museum-worthy, every chain seamless.

package/content/ARCHITECTURE.md

@@ -31,9 +31,11 @@ Modular system consisting of:
 │   ├── audio-design/        # Sound design rules
 │   └── prompt-structure/    # Prompt engineering patterns
 └── workflows/
-    ├── generate.md          # /generate command
-    ├── chain.md             # /chain command
-    └── safety-check.md      # /safety-check command
+    ├── generate.md          # /generate command
+    ├── chain.md             # /chain command
+    ├── safety-check.md      # /safety-check command
+    ├── recover.md           # /recover command
+    └── finish.md            # /finish command
 ```
 
 ---
@@ -64,10 +66,11 @@ Modular system consisting of:
 
 | Command | Description |
 |---------|-------------|
-| `/generate` | Generate shots and SAVE TO FILES |
-| `/chain` | Continue from last shot file |
-| `/safety-check` | Validate prompts before delivery |
-| `/finish` | Complete project, create summary |
+| `/generate` | Generate shots and SAVE TO FILES |
+| `/chain` | Continue from last shot file |
+| `/safety-check` | Validate prompts before delivery |
+| `/recover` | Recover failed safety/delivery gates |
+| `/finish` | Complete project, create summary |
 
 ---
 

package/content/FILM-KIT-INFO.md

@@ -30,10 +30,11 @@
 │   ├── audio-design/      ← Ses tasarımı
 │   └── prompt-structure/  ← Prompt şablonları
 └── workflows/             ← İş akışları
-    ├── generate.md        ← /generate komutu
-    ├── chain.md           ← /chain komutu
-    ├── safety-check.md    ← /safety-check komutu
-    └── finish.md          ← /finish komutu
+    ├── generate.md        ← /generate komutu
+    ├── chain.md           ← /chain komutu
+    ├── safety-check.md    ← /safety-check komutu
+    ├── recover.md         ← /recover komutu
+    └── finish.md          ← /finish komutu
 
 RULES.md                   ← AI editörler için kurallar (otomatik okunur)
 ```
@@ -44,10 +45,11 @@ RULES.md                   ← AI editörler için kurallar (otomatik okunur)
 
 | Komut | Ne Yapar |
 |-------|----------|
-| `/generate` | Senaryodan shot'lar üretir, dosyaya kaydeder |
-| `/chain` | Önceki shot'tan devam eder |
-| `/safety-check` | Tüm promptları güvenlik kontrolünden geçirir |
-| `/finish` | Projeyi tamamlar, özet oluşturur |
+| `/generate` | Senaryodan shot'lar üretir, dosyaya kaydeder |
+| `/chain` | Önceki shot'tan devam eder |
+| `/safety-check` | Tüm promptları güvenlik kontrolünden geçirir |
+| `/recover` | Hata/fail durumunda kurtarma akışı çalıştırır |
+| `/finish` | Projeyi tamamlar, özet oluşturur |
 
 ---
 

package/content/MASTER.md

@@ -48,7 +48,7 @@ Scenario Received → Check for elements:
 
 > 🚨 **CRITICAL:** Bu kuralı kullanıcının hatırlatmasını BEKLEME. PROAKTİF olarak uygula.
 
-**MUTLAK KURAL:** Senaryoda geçen TÜM gerçek kişi adlarını **otomatik olarak** fiziksel tanımlarla değiştir.
+**MUTLAK KURAL:** Senaryodaki gerçek kişi adlarını, **gorsel prompt ve metadata alanlarinda** otomatik fiziksel tanimlarla degistir.
 
 ```
 SENARYO GELİNCE:
@@ -581,10 +581,10 @@ Total: 1 main + [N] coverage = [N+1] production shots
 
 Before outputting, validate EVERY shot. **Bu kontrol otomatiktir, kullanıcı hatırlatması gerekmez.**
 
-### 1️⃣ AUTO-ANONYMOUS (EN ÖNCE)
-- [ ] TÜM isimler fiziksel tanımla değiştirildi mi? → **Otomatik yapılmalı**
-- [ ] Dolaylı ünlü atıfları temizlendi mi?
-- [ ] Tarihi kişi rolleri anonim mi?
+### 1️⃣ AUTO-ANONYMOUS (EN ÖNCE)
+- [ ] Gorsel prompt ve metadata icindeki tum isimler fiziksel tanimla degistirildi mi? -> **Otomatik yapilmali**
+- [ ] Dolaylı ünlü atıfları temizlendi mi?
+- [ ] Tarihi kişi rolleri anonim mi?
 
 ### 2️⃣ AUTO-SAFETY
 - [ ] Platform filtrelerine takılacak içerik var mı? → **Otomatik optimize et**

package/content/RULES.md

@@ -50,7 +50,8 @@ Film/Video request detected → Activate prompt-engineer agent
 | **Single shot** | "bir sahne", "one shot" | Generate 1 shot |
 | **Batch generation** | "3 shot", "5 sahne" | Generate N shots |
 | **Continue** | "devam et", "continue" | Chain from previous |
-| **Safety check** | /safety-check | Validate prompts |
+| **Safety check** | /safety-check | Validate prompts |
+| **Recover** | /recover | Recover failed safety/delivery gates |
 
 ---
 
@@ -207,10 +208,11 @@ Color: LUT, grade, desaturated, warm/cool, high/low key
 
 | Command | Purpose |
 |---------|--------|
-| `/generate` | Generate shots and SAVE TO FILES |
-| `/chain` | Continue from last shot file |
-| `/safety-check` | Validate before delivery |
-| `/finish` | Complete project, create summary |
+| `/generate` | Generate shots and SAVE TO FILES |
+| `/chain` | Continue from last shot file |
+| `/safety-check` | Validate before delivery |
+| `/recover` | Recover failed quality gates |
+| `/finish` | Complete project, create summary |
 
 ---
 

package/content/agents/prompt-engineer.md

@@ -29,7 +29,7 @@ You are a senior Technical Prompt Engineer specialized in Google Flow + Veo 3.1.
 2. **ENFORCE mandatory frame chaining for seamless visual continuity (CRITICAL)**
 3. **LOCK all characters/objects to user-provided reference images (CRITICAL)**
 4. **Maintain visual/audio continuity across all shots**
-4. **AUTO-ANONYMOUS: Otomatik olarak tüm isimleri fiziksel tanımlarla değiştir**
+4. **AUTO-ANONYMOUS:** Visual prompts and non-dialogue fields must stay anonymous
 5. **AUTO-SAFETY: Güvenlik filtrelerine takılacak içeriği proaktif reframe et**
 6. Apply dramatic pacing: "Ilmek ilmek işlemek" philosophy
 7. Use 8-second shot duration (default), slow burn, reaction shots
@@ -128,10 +128,11 @@ Embed this as physical behavior in the prompt, NOT as abstract metadata.
 
 Before outputting ANY shot:
 
-### 1. Safety & Anonymity (CRITICAL)
-- [ ] No real person's name? (Use physical descriptions)
-- [ ] No celebrity references?
-- [ ] No active violence/gore?
+### 1. Safety & Anonymity (CRITICAL)
+- [ ] No real person's name in visual prompts or metadata?
+- [ ] Dialogue naming follows `shot-plan.json -> dialogue_name_policy`?
+- [ ] No celebrity references?
+- [ ] No active violence/gore?
 
 ### 2. Prompt Structure
 - [ ] Follows 1-7 Flow Order?
@@ -156,9 +157,10 @@ For EACH shot, output exactly one file (`SHOTNN.md`) containing Main Shot + Cove
 
 ## Main Shot
 
-### İLK FRAME (SHOTNN_START)
-[If CHAINED: "→ Use SHOT[prev]_END as first frame"]
-[If FIRST/BREAK: Code block with Image Prompt (Flow Order + Avoid Line)]
+### İLK FRAME (SHOTNN_START)
+[Always provide a fenced code block]
+[If CHAINED: include "Use SHOT[prev]_END as exact first frame" inside code block]
+[If FIRST/BREAK: image prompt in same code block]
 
 ### SON FRAME (SHOTNN_END)
 ```