Skip to main content
Glama

MotionForge MCP

A local TypeScript MCP server for discovering product evidence, capturing real browser UI, directing evidence-bound timelines, exporting editable Remotion projects, rendering H.264 videos, and measuring technical export quality. No account or hosted infrastructure is required for the local workflow.

Version 0.1.0 implements 41 discoverable executable tools. Its creative tools produce structured artifacts from actual project evidence and client-authored copy; the AI client supplies judgment and validates product claims. It does not automatically guarantee a professional film or infer product functionality from a directory name.

Install

Prerequisites: Node.js >=20.19, npm, FFmpeg/FFprobe on PATH, and Chromium. Tested on macOS with Node20.20 and locally installed Google Chrome. Windows/Linux behavior is designed around portable Node APIs but has not been exercised here.

git clone https://github.com/2peng23/motionforge-mcp.git
cd motionforge-mcp
npm ci
npx playwright install chromium
npm run build
npm test

Remotion can use the same browser by setting MOTIONFORGE_BROWSER_PATH to the installed Chromium/Chrome executable; otherwise its renderer may need its own browser download. Preinstall a browser to keep production work offline.

MOTIONFORGE_ROOTS=/absolute/authorized/project MOTIONFORGE_BROWSER_PATH=/absolute/chrome node dist/server/cli.js

MOTIONFORGE_ROOTS is mandatory. Multiple roots use : on macOS/Linux, ; on Windows. Authorize both application source and output directory. The server reserves stdout for MCP messages; diagnostics and progress go to stderr. Future npm publishing is prepared through the executable bin entry; npx -y motionforge-mcp requires publishing first. Local package verification: npm pack --dry-run.

Related MCP server: voila

Connect a coding assistant

Client configurations cover Codex, Claude Code, Cursor and VS Code. SDK stdio discovery and tool invocation are integration-tested. Actual installation into those four apps was not performed and their configurations were not changed.

Run the acceptance demonstration

MOTIONFORGE_BROWSER_PATH="/Applications/Google Chrome.app/Contents/MacOS/Google Chrome" npm run demo

Omit the environment variable when Chromium is installed for both Playwright and Remotion. The demo starts an isolated localhost fixture, discovers its HTML pages, captures an actual working project filter and projects page, saves an evidence-bound storyboard, renders a 1920×1080 30fps 10-second MP4 with actual recorded interface footage, fully decodes it with FFmpeg, exports thumbnails and an editable Remotion project, and saves measured results under artifacts/demo-TIMESTAMP/acceptance-report.json. It stops its own fixture. No real customer data or supplied audio is used; the video is intentionally silent.

Typical MCP workflow

  1. analyze_project and detect_brand_identity inspect authorized source. Confirm feature claims through the running application.

  2. prepare_capture_environment reports development commands; it does not execute them. Obtain user authorization before starting the user's development server outside this MCP.

  3. capture_page and capture_user_flow capture actual routes. Mutating interactions require allowInteractions: true. Capture only accounts/data authorized for recording.

  4. The AI client authors a brief and shot titles from evidence. generate_storyboard saves the validated timeline object to an authorized new output path.

  5. create_motion_scene, animate_product_interface, create_camera_movement, and create_transition create new editable timeline versions.

  6. Optionally import licensed audio, mix it, and attach it with sync_audio_timeline.

  7. generate_remotion_project exports standalone source. render_video uses its own fixed trusted source, the validated timeline, and authorized local raster/video/audio assets.

  8. validate_export, inspect_audio_levels, analyze_rendered_frames, and generate_quality_report supply objective evidence for the client to inspect.

A timeline contains product, direction, width, height, fps, optional crf/encodingPreset/audio, and shots. Every shot requires title and evidence; asset is an authorized local raster image or MP4/WebM interaction clip; kind supports interface, typography, split, device and perspective; motion supports push, pan, reveal and hold; transition supports fade, wipe and cut. All exact schemas are discoverable via MCP tools/list. See workflow, tools, creative direction and security.

Scope and limits

  • Export: actual editable React/Remotion composition, manifest and copied local assets. Exported projects support npm install, npm run studio, npm run render.

  • Render: H.264 MP4/yuv420p, landscape/portrait/square or custom even dimensions, 12–60fps, up to 180 seconds and 30 scenes, CRF0–40, bounded encoding presets, one render at a time, configurable deadline. Portrait/square use their own layout rather than cropping landscape.

  • Browser: isolated contexts, fixed locale/timezone, responsive screenshots, scrolling/click/hover/form steps, cursor moves, source evidence and continuous masked recording (requested6–24fps, actual timing preserved; bounded30seconds/480frames/512MB). Private markers and fields are hidden; arbitrary personal images/text still require human review. See capture limitations in the returned evidence and security documentation.

  • Motion: deterministic frame-driven typography, camera pan/push, reveal masks, fade/wipe/cut, distinct screen/text/device layouts, and CSS perspective screen rotation. No arbitrary Three.js meshes, full 3D scene modeling, advanced layered parallax, or autonomous product storytelling is implemented. Device mockups are CSS frames, not photorealistic hardware.

  • Audio: authorized local audio import, FFmpeg mixing, timed client-authored scripts in any provided language (including English/Filipino/Taglish); no automatic translation. ElevenLabs is optional, requires explicit upload authorization and account credit; real paid synthesis was not tested. No stock-music downloader.

  • QA: actual FFprobe stream/duration/dimension checks, full decoding, black/freeze diagnostics, sampled frames and peak/mean audio metering. Text clipping measures actual Chrome text-range bounds and clipping ancestors at each scene midpoint (6% horizontal / 3% vertical safe area, 1px tolerance). It excludes text embedded in screenshots/video, all-frame coverage and complex clip-path guarantees; human frame review remains required. No computer-vision beauty score or complete aesthetic evaluation.

  • Framework discovery: dependency detection and heuristic route evidence, including React/Next, Vue/Nuxt, Svelte/SvelteKit and HTML. No framework migration, project script execution, or application-code rewrite.

The code is MIT licensed. Remotion has its own license and commercial eligibility rules; installing this MIT server does not grant a Remotion license. Optional provider terms and asset licenses also apply.

See verification evidence, troubleshooting, contributing, and API sources.

Available Tools

41 tools
analyze_projectC
Read-onlyIdempotent

Inspect bounded source inventory, framework, scripts, route evidence, assets and branding.

Inputs: project (required). Returns: result

analyze_rendered_framesB

Decode the video; run black/freeze detection and export sampled thumbnails for human review.

Inputs: input (required), outputDirectory (required). Returns: result

analyze_user_flowsB
Read-onlyIdempotent

Return actual route evidence for a client to author flows. Static scanning cannot verify runtime journeys.

Inputs: project (required). Returns: result

analyze_visual_rhythmB
Read-onlyIdempotent

Measure shot lengths, repetition and readability timing without inventing aesthetic scores.

Inputs: timeline (required). Returns: result

animate_product_interfaceC

Set implemented push, pan, reveal or hold choreography on a real timeline shot.

Inputs: output (required), timeline (required), shotIndex (required), motion. Returns: result

capture_pageB

Capture authentic screenshot in isolated browser; fields/private elements masked and third-party requests blocked.

Inputs: url (required), output (required), width, height, +8 more. Returns: result

capture_responsive_screensC

Capture an explicitly chosen responsive viewport matrix.

Inputs: url (required), outputDirectory (required), height, widths. Returns: result

capture_user_flowC

Execute explicitly authorized bounded actions and capture the final real page state.

Inputs: url (required), output (required), width, height, +8 more. Returns: result

configure_renderB

Validate and save dimensions, frame rate, duration and encoding configuration as a renderable timeline.

Inputs: output (required), timeline (required). Returns: result

create_camera_movementC

Apply deterministic camera motion to an existing scene.

Inputs: output (required), timeline (required), shotIndex (required), motion. Returns: result

create_motion_sceneB

Validate and append a real shot to a new timeline file; no destructive overwrite.

Inputs: shot (required), output (required), timeline (required). Returns: result

create_scene_planB
Read-onlyIdempotent

Calculate exact scene frame ranges from a validated timeline.

Inputs: timeline (required). Returns: result

create_transitionC

Apply a supported fade, wipe or cut transition in a new timeline.

Inputs: output (required), timeline (required), shotIndex (required), transition. Returns: result

create_typography_animationC

Create frame-driven typography reveal on a selected shot.

Inputs: output (required), timeline (required), shotIndex (required). Returns: result

detect_brand_identityB
Read-onlyIdempotent

Extract existing source colors, font declarations and real headings without inventing branding.

Inputs: project (required). Returns: result

detect_clippingA

Measure actual composition text bounds and clipping ancestors in Chrome at each scene midpoint; excludes text inside screenshots/video. Temporary QA files are cleaned up.

Inputs: timeline (required). Returns: result

discover_routesB
Read-onlyIdempotent

Find filesystem route candidates with source paths and explicit inference confidence.

Inputs: project (required). Returns: result

export_projectC

Export a standalone editable Remotion project with source assets and npm manifest.

Inputs: timeline (required), outputDirectory (required). Returns: result

export_videoC

Validate the complete video stream before copying a local MP4 artifact.

Inputs: input (required), output (required). Returns: result

generate_3d_sceneC

Apply deterministic CSS perspective rotation to a product screen. No mesh/Three.js scene generation.

Inputs: output (required), timeline (required), shotIndex (required). Returns: result

generate_creative_briefC
Read-onlyIdempotent

Build an evidence-grounded creative brief from source identity, actual routes and explicit audience/goal.

Inputs: goal (required), project (required), audience (required), language, +1 more. Returns: result

generate_device_mockupC

Frame an actual screenshot with an animated device border; no stock device download.

Inputs: output (required), timeline (required), shotIndex (required). Returns: result

generate_quality_reportD

Persist measured decode/stream/audio checks and explicit manual review requirements.

Inputs: input (required), output (required). Returns: result

generate_remotion_projectC

Export actual editable React/Remotion source, validated timeline and local assets.

Inputs: timeline (required), outputDirectory (required). Returns: result

generate_storyboardC

Validate evidence-bound shots and persist a real renderable timeline. Client supplies product claims and screenshots.

Inputs: output (required), timeline (required). Returns: result

generate_voiceoverC

Optionally synthesize supplied text with ElevenLabs. Explicit upload permission and paid account credentials required.

Inputs: text (required), output (required), voiceId (required), authorizeUpload. Returns: result

generate_voiceover_scriptB
Read-onlyIdempotent

Create a timed script from client-authored evidence-bound scene copy; preserves provided language.

Inputs: timeline (required). Returns: result

import_audioC

Copy a local authorized music/voice track with license provenance.

Inputs: input (required), output (required), license (required). Returns: result

inspect_animation_opportunitiesB

Inspect actual headings, links, overflow and Web Animations timing for motion planning; writes evidence screenshot.

Inputs: url (required), output (required), width, height, +8 more. Returns: result

inspect_audio_levelsB
Read-onlyIdempotent

Measure actual FFmpeg peak/mean levels and report clipping risk.

Inputs: input (required). Returns: result

inspect_media_assetsB
Read-onlyIdempotent

Inventory local image, font, video and audio assets and byte sizes.

Inputs: project (required). Returns: result

inspect_renderC
Read-onlyIdempotent

Read actual FFprobe streams and container metadata.

Inputs: input (required). Returns: result

mix_audioC

Mix real local tracks with bounded gain/delay and output limiter via FFmpeg.

Inputs: output (required), tracks (required), duration (required). Returns: result

plan_sound_designB
Read-onlyIdempotent

Calculate scene boundaries and audio cue windows from actual timeline without downloading music.

Inputs: timeline (required). Returns: result

plan_transitionsB
Read-onlyIdempotent

Return implemented transition types with their timeline positions; no random transition assignment.

Inputs: timeline (required). Returns: result

prepare_capture_environmentB
Read-onlyIdempotent

Check project commands and capture URL; does not execute arbitrary project scripts or start a server.

Inputs: project (required). Returns: result

record_browser_interactionC

Record real browser interactions through continuously masked screenshots, cursor choreography and measured frame timing; exports local MP4.

Inputs: url (required), output (required), width, height, +8 more. Returns: result

render_videoC

Render a bounded real H.264 MP4 with Remotion and retain editable project and progress log.

Inputs: output (required), timeline (required). Returns: result

sync_audio_timelineC

Attach an authorized local audio file to an exported timeline.

Inputs: audio (required), output (required), timeline (required). Returns: result

validate_exportB
Read-onlyIdempotent

Fully decode exported media and verify codec and optional dimensions.

Inputs: input (required), fps, width, height, +1 more. Returns: result

validate_timelineA
Read-onlyIdempotent

Validate schema, timings, local assets and bounds before rendering.

Inputs: timeline (required). Returns: result

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    C2.7/5.0

    Scored across 41 tools

    Disambiguation2/5

    Several clusters have unclear boundaries despite verbose descriptions: render_video vs export_video vs export_project vs generate_remotion_project all appear to produce artifacts from a timeline, and export_project vs generate_remotion_project read as near-duplicates. Similarly validate_timeline / validate_export / inspect_render / analyze_rendered_frames form an overlapping QA/validation cluster, and generate_storyboard / create_scene_plan / create_motion_scene / generate_remotion_project are four plausible entry points for 'build a timeline'. An agent would need to read full descriptions to avoid misselection.

    Naming Consistency5/5

    Every tool follows a snake_case verb_noun pattern (import_audio, capture_page, validate_export, create_transition, etc.), with verbs drawn from a coherent vocabulary across the pipeline stages. The only nit is near-identical names like generate_voiceover vs generate_voiceover_script, but the structural convention itself is uniform.

    Tool Count2/5

    41 tools is well above the comfortable range for a single server and the surface reads as several sub-servers bolted together (project analysis, browser capture, timeline authoring, audio, rendering, QA/export). Many tools are thin single-step operations that could be consolidated into a handful of parameterized tools.

    Completeness4/5

    The lifecycle is unusually thorough: analysis, evidence capture, creative planning, timeline authoring, rendering, validation, QA reporting and export are all represented. Gaps are minor but real — no tools to update/reorder/remove existing scenes or shots, no project listing, and no subtitle/text-overlay burn-in step.

    Maintenance

    ActivityMaintained
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      B
      maintenance
      Enables AI agents to create polished product demo videos by controlling a real Chromium browser, recording actions, and rendering 1080p MP4s with narration, captions, and styled overlays.
      2
      MIT
    • A
      license
      Not graded
      quality
      B
      maintenance
      Enables agents to record narrated, zoom-animated product demo videos of websites entirely on-device, including outlining pages, scripting tours, recording MP4s, reviewing frames, and extracting embedded recipes to fork or rerender demos.
      38 npm
      6
      MIT
    • A
      license
      B
      quality
      F
      maintenance
      Enables an agent to drive and record macOS apps with real clicks, typing and taps, then analyze the footage and edit it live in the screen recorder's editor with cuts, speed-ups, zooms, narration, captions and music, and export sized variants. Supports seven editing recipes, brand kits, privacy blurring, undo-safe checkpoints, and use from Claude Code or Codex.
      46
      54 npm
      MIT