Rebuild MoneyPrinterTurbo as a no-key AI video studio

.github/workflows/ci.yml

@@ -0,0 +1,22 @@
+name: CI
+
+on:
+  push:
+    branches: [main, "agent/**"]
+  pull_request:
+
+jobs:
+  test:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.11"
+          cache: pip
+      - name: Install
+        run: python -m pip install -e ".[dev]"
+      - name: Compile
+        run: python -m compileall -q app
+      - name: Test
+        run: pytest -q

.gitignore

@@ -0,0 +1,10 @@
+__pycache__/
+*.py[cod]
+.venv/
+.env
+.DS_Store
+.pytest_cache/
+.ruff_cache/
+storage/*
+!storage/.gitkeep
+*.log

Dockerfile

@@ -0,0 +1,25 @@
+FROM node:22-bookworm-slim
+
+ENV PYTHONDONTWRITEBYTECODE=1 \
+    PYTHONUNBUFFERED=1 \
+    VIRTUAL_ENV=/opt/venv \
+    PATH="/opt/venv/bin:$PATH"
+
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    python3 python3-venv python3-pip ffmpeg ca-certificates \
+    && rm -rf /var/lib/apt/lists/*
+
+RUN python3 -m venv /opt/venv
+RUN npm install -g @openai/codex @anthropic-ai/claude-code @google/gemini-cli
+
+WORKDIR /app
+COPY pyproject.toml README.md LICENSE ./
+COPY app ./app
+COPY web ./web
+RUN pip install --no-cache-dir .
+
+RUN mkdir -p /app/storage/uploads /app/storage/tasks
+VOLUME ["/app/storage"]
+EXPOSE 8501
+
+CMD ["uvicorn", "app.main:app", "--host", "0.0.0.0", "--port", "8501"]

LICENSE

@@ -0,0 +1,22 @@
+MIT License
+
+Copyright (c) 2024 Harry
+Copyright (c) 2026 GenRamzi contributors
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

NOTICE.md

@@ -0,0 +1,11 @@
+# Attribution and project notice
+
+MoneyPrinterTurbo NoAPI is a modified/rebuilt project inspired by and derived from ideas and portions of the MIT-licensed MoneyPrinterTurbo project:
+
+- Upstream: https://github.com/harry0703/MoneyPrinterTurbo
+- Upstream author/copyright notice: Copyright (c) 2024 Harry
+- Upstream license: MIT
+
+This repository adds a new account/CLI provider architecture, a new web interface, a local-first media workflow, new task orchestration, and NoAPI-specific documentation and behavior.
+
+OpenAI, ChatGPT, Codex, Anthropic, Claude, Google, Gemini, Ollama, and other product names are trademarks of their respective owners. This project is not endorsed by or affiliated with those vendors.

README.md

@@ -1,3 +1,161 @@
 # MoneyPrinterTurbo NoAPI
 
-Modern no-key rebuild of MoneyPrinterTurbo. Development happens on feature branches and is merged through pull requests.
+A modern, local-first rebuild of [MoneyPrinterTurbo](https://github.com/harry0703/MoneyPrinterTurbo) that removes the old **LLM Settings / API-key workflow** and replaces it with **connected AI accounts and local models**.
+
+> Status: first functional rebuild. The UI, account-provider layer, local media pipeline, TTS, subtitles, task runner, and FFmpeg renderer are implemented. This repository intentionally does **not** scrape browser cookies or impersonate ChatGPT/Claude/Gemini web sessions.
+
+## What changed
+
+- New responsive web studio instead of the large Streamlit settings screen.
+- No user API-key form.
+- ChatGPT through the official **Codex CLI** after `codex login`.
+- Claude through the official **Claude Code CLI** after account login.
+- Gemini through the official **Gemini CLI** after Google login.
+- Ollama as a fully local fallback with no cloud account.
+- Prompt → script → neural TTS → SRT subtitles → media composition → MP4 pipeline.
+- Local image/video uploads, background music, 9:16 / 16:9 / 1:1 output, and batch generation.
+- FastAPI backend with task progress and downloadable results.
+- No credential copying: account credentials stay where the official provider CLI stores them.
+
+## Important meaning of “NoAPI”
+
+“NoAPI” means **the user does not enter or manage an API key in this application**. Cloud models still communicate with their vendors through the vendor’s official authenticated software. ChatGPT, Claude, and Gemini subscriptions/usage rules still apply. Ollama is the option for fully local inference.
+
+The app does not use unofficial cookie extraction, hidden browser endpoints, session-token copying, or credential harvesting.
+
+## Requirements
+
+- Python 3.11+
+- FFmpeg + ffprobe available in `PATH`
+- At least one AI provider below
+- Internet access for cloud-account providers and Edge neural TTS
+
+### ChatGPT
+
+```bash
+npm install -g @openai/codex
+codex login
+codex login status
+```
+
+### Gemini
+
+```bash
+npm install -g @google/gemini-cli
+gemini
+```
+
+Inside Gemini CLI, choose **Login with Google**. After that, MoneyPrinterTurbo NoAPI uses Gemini CLI headless mode for script generation.
+
+### Claude
+
+```bash
+npm install -g @anthropic-ai/claude-code
+claude auth login
+claude auth status
+```
+
+### Ollama (fully local)
+
+Install Ollama, then:
+
+```bash
+ollama pull qwen3:8b
+ollama list
+```
+
+Change the default local model with:
+
+```bash
+export MPT_OLLAMA_MODEL="qwen3:8b"
+```
+
+On Windows PowerShell:
+
+```powershell
+$env:MPT_OLLAMA_MODEL="qwen3:8b"
+```
+
+## Run locally
+
+```bash
+git clone https://github.com/GenRamzi/MoneyPrinterTurboNoAPI.git
+cd MoneyPrinterTurboNoAPI
+python -m venv .venv
+```
+
+Activate the virtual environment, then:
+
+```bash
+pip install -e .
+python main.py
+```
+
+Open `http://127.0.0.1:8501`.
+
+The UI can launch a provider login terminal on common desktop systems. If the app is running in Docker, a remote server, or a desktop environment where terminal launching is unavailable, it shows the exact login command instead.
+
+## Docker
+
+Docker is useful for the renderer and Ollama-oriented deployments. Account-based CLI login is usually simpler when running the app directly on the host OS.
+
+```bash
+docker compose up --build
+```
+
+Then open `http://localhost:8501`.
+
+## Current video workflow
+
+1. Choose a connected AI account.
+2. Enter a topic or paste your own script.
+3. Choose language, duration, aspect ratio, voice, and subtitle position.
+4. Upload any number of images/videos; they are normalized and rotated across batch outputs.
+5. Optionally add background music.
+6. Generate. The task engine writes the script, voice, captions, intermediate media, and final MP4 files under `storage/tasks/<task-id>/`.
+
+If no visual media is uploaded, the renderer creates a clean generated background so the pipeline remains usable without any stock-media API.
+
+## Architecture
+
+```text
+web/                         Modern zero-build frontend
+app/main.py                  FastAPI entrypoint + SPA hosting
+app/api.py                   Providers, uploads, tasks, downloads
+app/providers/               Official CLI account adapters
+app/services/script.py       AI script generation
+app/services/tts.py          Edge neural TTS
+app/services/subtitles.py    SRT timing/generation
+app/services/media.py        FFmpeg normalization/composition
+app/services/tasks.py        Background job orchestration
+storage/                     User uploads and generated projects
+```
+
+Provider adapters are intentionally isolated. Adding another provider only requires implementing the `TextProvider` interface; the creator UI consumes the registry automatically.
+
+## Security decisions
+
+- No API keys are accepted by the web UI.
+- No browser cookies are requested, parsed, or stored.
+- Login commands are fixed per provider; user input is never concatenated into an arbitrary shell command.
+- Upload and task-download paths are normalized and constrained to application storage.
+- Provider credentials remain managed by the official CLIs.
+- Generated files are local unless the user deploys this server elsewhere.
+
+## Differences from upstream
+
+This is a clean architectural rebuild inspired by and derived from MoneyPrinterTurbo’s MIT-licensed concept and workflow. It is not a claim of affiliation with Harry or the AI vendors. See `NOTICE.md` and `LICENSE`.
+
+The original project supports many API-backed LLM and stock-media providers. This rebuild deliberately prioritizes **account login, local inference, local materials, and a much simpler product surface** instead of reproducing API-key configuration screens.
+
+## Development
+
+```bash
+pip install -e ".[dev]"
+pytest -q
+python -m compileall -q app
+```
+
+## License
+
+MIT. Upstream copyright and license notice are retained as required by the original license.

app/__init__.py

@@ -0,0 +1 @@
+__version__ = "0.1.0"

app/api.py

@@ -0,0 +1,104 @@
+from __future__ import annotations
+
+import shutil
+import uuid
+from pathlib import Path
+
+from fastapi import APIRouter, File, HTTPException, UploadFile
+from fastapi.responses import FileResponse
+
+from app.core.config import settings
+from app.models.schemas import GenerateRequest, TaskInfo
+from app.providers.registry import provider_registry
+from app.services.media import AUDIO_EXT, IMAGE_EXT, VIDEO_EXT
+from app.services.tasks import task_manager
+
+router = APIRouter(prefix="/api")
+ALLOWED_UPLOADS = VIDEO_EXT | IMAGE_EXT | AUDIO_EXT
+
+
+@router.get("/health")
+def health() -> dict:
+    return {
+        "ok": True,
+        "app": settings.app_name,
+        "ffmpeg": bool(shutil.which(settings.ffmpeg_bin)),
+        "ffprobe": bool(shutil.which(settings.ffprobe_bin)),
+    }
+
+
+@router.get("/providers")
+def providers():
+    return provider_registry.list()
+
+
+@router.post("/providers/{provider_id}/login")
+def provider_login(provider_id: str) -> dict:
+    try:
+        message = provider_registry.open_login(provider_id)
+    except Exception as exc:
+        raise HTTPException(status_code=404, detail=str(exc)) from exc
+    return {"ok": True, "message": message}
+
+
+@router.get("/voices")
+def voices() -> list[dict[str, str]]:
+    return [
+        {"id": "ar-SA-HamedNeural", "name": "Hamed — Arabic (Saudi)"},
+        {"id": "ar-SA-ZariyahNeural", "name": "Zariyah — Arabic (Saudi)"},
+        {"id": "ar-EG-ShakirNeural", "name": "Shakir — Arabic (Egypt)"},
+        {"id": "ar-EG-SalmaNeural", "name": "Salma — Arabic (Egypt)"},
+        {"id": "en-US-AndrewNeural", "name": "Andrew — English (US)"},
+        {"id": "en-US-AvaNeural", "name": "Ava — English (US)"},
+        {"id": "en-GB-RyanNeural", "name": "Ryan — English (UK)"},
+        {"id": "en-GB-SoniaNeural", "name": "Sonia — English (UK)"},
+    ]
+
+
+@router.post("/uploads")
+async def upload(files: list[UploadFile] = File(...)) -> list[dict[str, str]]:
+    saved: list[dict[str, str]] = []
+    max_bytes = settings.max_upload_mb * 1024 * 1024
+    for file in files:
+        suffix = Path(file.filename or "").suffix.lower()
+        if suffix not in ALLOWED_UPLOADS:
+            raise HTTPException(status_code=415, detail=f"Unsupported file type: {suffix or 'unknown'}")
+        data = await file.read(max_bytes + 1)
+        if len(data) > max_bytes:
+            raise HTTPException(status_code=413, detail=f"File exceeds {settings.max_upload_mb} MB")
+        file_id = f"{uuid.uuid4().hex}{suffix}"
+        target = settings.uploads_dir / file_id
+        target.write_bytes(data)
+        saved.append({"id": file_id, "name": file.filename or file_id})
+    return saved
+
+
+@router.post("/tasks", response_model=TaskInfo)
+def create_task(request: GenerateRequest) -> TaskInfo:
+    return task_manager.create(request)
+
+
+@router.get("/tasks", response_model=list[TaskInfo])
+def list_tasks() -> list[TaskInfo]:
+    return task_manager.list()
+
+
+@router.get("/tasks/{task_id}", response_model=TaskInfo)
+def get_task(task_id: str) -> TaskInfo:
+    task = task_manager.get(task_id)
+    if not task:
+        raise HTTPException(status_code=404, detail="Task not found")
+    return task
+
+
+@router.get("/tasks/{task_id}/files/{filename}")
+def task_file(task_id: str, filename: str):
+    if Path(task_id).name != task_id or Path(filename).name != filename:
+        raise HTTPException(status_code=400, detail="Invalid path")
+    task = task_manager.get(task_id)
+    if not task or filename not in task.output_files:
+        raise HTTPException(status_code=404, detail="File not found")
+    path = (settings.tasks_dir / task_id / filename).resolve()
+    if path.parent != (settings.tasks_dir / task_id).resolve() or not path.exists():
+        raise HTTPException(status_code=404, detail="File not found")
+    return FileResponse(path, media_type="video/mp4", filename=filename)