OpenMontage

Delegates video production orchestration to the LLM running in the user's IDE (Claude Code, Cursor, Windsurf) rather than making runtime API calls for control logic. The agent reads YAML pipeline manifests, interprets specialized skill instructions, executes Python tools sequentially, and persists state via checkpoint files. This eliminates latency and cost of cloud orchestration while keeping the user's coding assistant as the control plane.

Structures all video production work into YAML-defined pipeline stages with explicit inputs, outputs, and tool sequences. Each pipeline manifest declares a series of named stages (e.g., 'script', 'asset_generation', 'composition') with tool dependencies and human approval gates. The agent reads these manifests to understand the production flow and enforces 'Rule Zero' — all production requests must flow through a registered pipeline, preventing ad-hoc execution.

Provides a pipeline for generating talking head videos where a digital avatar or real person speaks a script. The system supports multiple avatar providers (D-ID, Synthesia, Runway), voice cloning for consistent narration, and lip-sync synchronization. The agent can generate talking head videos from text scripts without requiring video recording or manual editing.

Provides a pipeline for generating cinematic videos with planned shot sequences, camera movements, and visual effects. The system includes a shot prompt builder that generates detailed cinematography prompts based on shot type (wide, close-up, tracking, etc.), lighting (golden hour, dramatic, soft), and composition principles. The agent orchestrates image generation, video composition, and effects to create cinematic sequences.

Provides a pipeline for converting long-form podcast audio into short-form video clips (TikTok, YouTube Shorts, Instagram Reels). The system extracts key moments from podcast transcripts, generates visual assets (images, animations, text overlays), and creates short videos with captions and background visuals. The agent can repurpose a 1-hour podcast into 10-20 short clips automatically.

Provides an end-to-end localization pipeline that translates video scripts to multiple languages, generates localized narration with native-speaker voices, and re-composes videos with localized text overlays. The system maintains visual consistency across language versions while adapting text and narration. A single source video can be automatically localized to 20+ languages without re-recording or re-shooting.

Implements a tool registry system where all video production tools (image generation, TTS, video composition, etc.) inherit from a BaseTool contract that defines a standard interface (execute, validate_inputs, estimate_cost). The registry auto-discovers tools at runtime and exposes them to the agent through a standardized API. This allows new tools to be added without modifying the core system.

Implements Meta Skills that enforce quality standards and production governance throughout the pipeline. This includes human approval gates at critical stages (after scripting, before expensive asset generation), quality checks (image coherence, audio sync, video duration), and rollback mechanisms if quality thresholds are not met. The system can halt production if quality metrics fall below acceptable levels.

Provides a pipeline for generating screen recording videos and software demo videos. The system can capture screen recordings, add narration and captions, highlight UI elements, and create polished demo videos. The agent can generate demo videos from descriptions of software features without requiring manual screen recording or editing.

Implements a provider selector pattern where every video generation, image generation, and audio capability supports both high-end cloud APIs (OpenAI, Anthropic, ElevenLabs, Runway) and local/open-source alternatives (Stable Diffusion, Ollama, FFmpeg). The system scores available providers based on cost, latency, quality, and GPU availability, then selects the best match for the current task. This allows users to start with free local models and upgrade to premium APIs without code changes.

Provides specialized instruction sets ('skills') that teach the agent how to execute specific production tasks (e.g., 'Cinematic Rendering', 'Talking Head Generation', 'Podcast Repurposing'). Skills are organized into Core Skills (foundational operations), Creative Skills (style and composition), and Meta Skills (governance and quality). Each skill contains detailed prompts, examples, and decision trees that guide the agent through complex multi-step processes without requiring the agent to invent the approach.

Provides a Remotion-based composition engine that generates videos from declarative JSON scene definitions. The system includes pre-built Remotion components for explainers, cinematic renders, talking heads, and animated sequences. The agent can generate or modify Remotion composition JSON, which is then rendered via the Remotion CLI to produce final video output. This enables programmatic video generation without manual editing.

Provides multi-provider TTS capabilities supporting both cloud APIs (ElevenLabs, OpenAI, Google Cloud) and local alternatives (Ollama, Coqui). Supports voice cloning for consistent narrator voices across videos, automatic language detection and translation for localization, and voice profile management. The system can generate speech in 50+ languages and apply voice effects (speed, pitch, emotion) without re-recording.

Provides image generation capabilities (Flux, Stable Diffusion, DALL-E) with a built-in style system and cinematography framework. The system includes pre-defined style playbooks (cinematic, documentary, animated, etc.) and a shot prompt builder that generates detailed image prompts based on cinematography principles (framing, lighting, composition). The agent can apply consistent visual styles across multiple images without manually crafting detailed prompts.

Tracks API costs across all provider calls (OpenAI, Anthropic, ElevenLabs, Runway, etc.) in real-time and enforces budget limits per pipeline or per production. The system logs cost per tool execution, aggregates costs by provider and pipeline stage, and can halt expensive operations if budget is exceeded. Provides cost estimates before executing expensive operations (e.g., video generation) to enable informed decision-making.

Implements a checkpoint system that saves the state of each pipeline stage to JSON files, enabling resumption of interrupted productions without re-executing completed stages. Each checkpoint includes stage outputs, tool execution logs, and metadata (timestamp, cost, quality metrics). If a pipeline fails mid-execution, the agent can resume from the last checkpoint, skipping already-completed stages and re-executing only the failed stage.

Provides an end-to-end pipeline for generating animated explainer videos from text descriptions. The pipeline includes script generation, scene breakdown, image generation for each scene, text-to-speech narration, and Remotion-based composition. The agent follows the Animated Explainer skill to create visually coherent, well-paced explainer videos with synchronized narration and animations.