MCP App Store
by KeenEthics
Rendley icon

Rendley MCP Server for Claude and ChatGPT

By Rendley

Overview

Create and edit videos with prompts. Rendley can use your footage, files, links, and other context to cut clips, repurpose content, resize videos for different formats, add captions, or create something entirely new. Then open the project in the editor and fine-tune anything you want.

How to connect

  1. Claude: Open the Rendley listing in the Claude connectors directory and choose Connect. Claude listing Rendley exposes 8 tools here.
  2. ChatGPT: Open the Rendley listing in the ChatGPT app directory and choose Connect. ChatGPT listing Rendley exposes 24 tools here.

Tools

add_files

ChatGPT
Adds files to a project or to the workspace library. files: files the user attached in this chat: pass each one's path (e.g. /mnt/data/clip.mp4) and ChatGPT hands the file over. urls for links (they import in the background), or library_assets to copy library assets into the project. With only project_id, it returns a page where the user adds files from their own device; use that when the files are on the user's computer or phone and you can't reach them. Pass the original files: Rendley converts, trims and resizes inside the project, so nothing needs converting or compressing first (video, audio and images up to 5 GB each). Never invent URLs such as attachment:// or re-host files elsewhere.

create_project

ChatGPT
Creates a video project: empty, or from a template with its fields filled in (read them with get_template first). Add media with add_files, then build it with edit_project or run_agent.

create_template

ChatGPT
Saves a copy of a project, with its files, as a template of its workspace, exposing the given properties as fields: texts, colors, numbers, the files clips play. Read the project with get_project first for clip and media ids. Calling it again for the same project updates that template. Keyframed values can't be fields.

create_workspace

ChatGPT
Creates a workspace for organizing projects, media and a brand kit. It counts toward the plan's workspace limit.

delete_project

ChatGPT
Permanently deletes a project and everything in it. This can't be undone. Only use it when the user asked to delete this specific project.

duplicate_project

ChatGPT
Copies a project with its media, for variants such as a vertical version. The original is unchanged.

edit_project

ChatGPT
Applies changes to a project's timeline as one atomic save: trim, cut, split, move, speed, volume, fades, text and titles, transitions, filters, color, keyframes, crop, layout, canvas size and caption style. Every kind of change and its fields are in this tool's input schema. Uses no credits. Returns each change's result and the timeline's warnings; include_timeline adds the project overview. Format: {"commands": [{"cmd": "moveClip", "args": {...}, "rets": ["clipA"]}, ...]}, one entry per change, cmd naming its kind. A change can name what it returns in rets; later changes in the same save reference it in args as ${name} (a list return as ${name[0]}). Times are seconds; boxes are canvas pixels from the top-left. The changes save together: if any fails, nothing is saved. A repeat of a saved call (same request_id, by default the same program) within 10 minutes is answered without applying it again. Captions from speech and cutting filler words need a transcription: ask run_agent for them.

export_video

ChatGPT
Cloud export: renders the project into a video file on Rendley's servers and returns a download link. It needs a Rendley plan that includes cloud export and uses the same credits as cloud export in the app. Returns a job; follow it with get_job.

get_account

ChatGPT
Returns the user's own plan and status, credit balance, whether cloud export is included, storage, and project, workspace and transcription usage against their limits. Use it when the user asks about their plan, credits or usage, or when a tool reports a plan or credit limit.

get_brandkit

ChatGPT
Returns a workspace's brand voice, colors, fonts, text styles, caption style, voices, and reusable media assets.

get_job

ChatGPT
Returns the state of a run_agent run or an export: running, paused for the user's decision, finished with its result, or failed. The request may wait briefly for a change before returning.

get_job_progress

ChatGPT
Refreshes the job or upload status card shown to the user. Called by the status card itself.

get_media

ChatGPT
Returns one of the user's files by media_id, with a temporary download link: a file in a project (pass project_id), media Rendley's agent generated included, or a file in the workspace library. Use it to give the user a file or to work with it yourself. To put a library file into another project, use add_files library_assets, not this link.

get_motion_clip

ChatGPT
Returns a motion clip's document: the JSON that describes what it draws and how it moves (its scenes, elements, positions, sizes, colors, text and timing). Use it to change a motion clip (fix an overlap or a cut-off label, restyle it, or redesign it), then save the new document with edit_project's setMotionClipScript; get_reference topic motion_clip has the format. Read-only.

get_project

ChatGPT
Reads a project: its details, editor link, the files added to it (with their media_id) and, by default, an overview of every layer and clip with the timeline's warnings (text off screen, overlaps, sound playing twice). Read it before editing a project you haven't seen in this conversation. For more, pass paths: one clip's properties, a time range, captions, media. Nothing is rendered or changed. Paths (ids are real SDK uuids taken from earlier results): project summary: display, fps, counts, timelineEnd, selection project/display {width, height} selection currently selected clip ids layers every layer, bottom to top, with clip counts layers/<layerId>/clips the clips on one layer clips every timeline clip, in play order (id, type, layerId, startTime, endTime, duration, mediaName) clips/<clipId> one clip with all of its properties clips/<clipId>/<field> one property — startTime, endTime, duration, sourceDuration, leftTrim, rightTrim, text, layerId, rotation, position, scale, alpha, zIndex, crop, zoom, keyframes, effects, filters, style clips/<clipId>/media the library media that clip plays media library media summaries, with name and customData media/<mediaId>/<field> name, duration, width, height, type, filename, hash, mimeType, status, customData (customData: what the file holds — description is the prompt that generated it, url its origin; status: the editor's load state — loading/transcoding means still loading, error means the browser could not open the file) subtitles caption tracks with their lines and word times transitions every transition with its layer Filters on collections: clips?type=text, clips?layer=<id>, clips?media=<id>, clips?at=12.5, clips?from=0&to=10, media?type=video Timing is in the commands' own terms: startTime is where a clip starts playing (what setClipStartTime takes), endTime where it stops, duration how long it plays; keyframe times count from startTime; sourceDuration is the untrimmed source block. Each layer in layers lists its gaps (where it shows nothing). To change how long a clip shows, setTrimDuration. Transforms (rotation in radians, position [x,y], scale [x,y], alpha) are stored in the clip style — ask for them by the plain names listed above.

get_reference

ChatGPT
Looks up the catalog ids a project's editor offers (titles, transitions, filters, effects, shapes, lower thirds, blocks) for edit_project, details on edit_project's changes (return values and usage notes), and the motion clip document format for designing animated graphics. Read-only.

get_template

ChatGPT
Lists the fields a template exposes, each with the key create_project's fields take, its type (text, number, color, boolean, list with its length, or media with its media_kind) and the value the template ships with. Call it before create_project with template_id so the values are addressed by the right keys.

list_library

ChatGPT
Lists the assets in a workspace's media library (logos, b-roll, music the user saved) with names, types and tags. Bring one into a project with add_files' library_assets.

list_projects

ChatGPT
Lists the projects in a workspace, newest first. Use it when the user wants to work on an existing project and hasn't said which one.

list_templates

ChatGPT
Lists the templates saved in the user's workspace. A template is a project someone prepared for reuse, with the parts meant to change exposed as fields. Use it when the user asks to reuse a template, mentions one by name, or wants to see what templates they have. Follow with get_template to read the fields of the one they pick.

list_workspaces

ChatGPT
Lists the user's Rendley workspaces by name. A workspace holds projects, a media library and a brand kit.

preview_frames

ChatGPT
Renders what the project looks like at given times on Rendley's servers, as the viewer would see it (video, text, titles, captions, transitions and effects), and returns the image. Each call uses 1 credit from the user's balance, refunded if it fails. One time gives one frame; up to 36 give one contact sheet whose tiles are numbered, so you and the user can refer to a moment by its number ("#3"). Takes several seconds and changes nothing in the project.

run_agent

ChatGPT
Starts Rendley's editing agent on a request described in plain language (for example 'turn this interview into a 45 second vertical clip with captions'), or continues a paused run. Use it when the user wants Rendley to plan and carry out an edit. The agent plans the work, shows the plan with its credit price, and pauses for the user's approval before spending credits. Follow the run with get_job; answer a pause by calling run_agent with job_id and response.

update_brandkit

ChatGPT
Updates a workspace's brand kit: fill it from the brand's website, add colors, add logos, font files and other brand files (from a URL or the library), or set the brand voice. Use it when the user asks to set up or change their brand. Filling from a website can replace the kit's logo, colors, fonts and voice, and brand_voice replaces the current voice.

add_brand_colors

Claude

add_files

Claude

create_video_draft

Claude

get_brandkit

Claude

get_project

Claude

list_projects

Claude

list_workspaces

Claude

plan_video

Claude

Similar Apps

Listed on MCP App Store

Listed on MCP App Store

Add this badge to your site or README. It links to this page.

HTML

<a href="https://mcp-app-store.keenethics-labs.com/app/rendley"><img src="https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg" alt="Listed on MCP App Store" width="220" height="44"></a>

Markdown

[![Listed on MCP App Store](https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg)](https://mcp-app-store.keenethics-labs.com/app/rendley)