Integration

Abyssale integration

Abyssale is a creative automation platform that renders marketing images, videos, animated GIFs, HTML5 banners and print PDFs from designs built in its editor. An agent can list and inspect designs, projects, fonts and workspace templates, generate one image synchronously or many formats asynchronously (then read the finished files), export generated files as a ZIP, create a dynamic image URL, rename, move or delete designs, duplicate a workspace template into a project, create designs from JSON (alpha) and read the remaining credits. Generation spends the workspace's credits. The API key check route and the webhook signing secret routes are not exposed. Authentication is an API key from the workspace settings, sent in the x-api-key header.

Connect Abyssale How integrations work

Ask it in chat

Say what should happen.
The workflow builds itself.

When this happens, do that in Abyssale.Built from the 25 Abyssale tools below, on the connection you add once.

What you can do with Abyssale

Each of these is one operation on one endpoint, available both as a workflow step and as a tool an AI agent can call on its own. Nothing bundles several operations into one action, so what a run did stays readable afterwards.

Most endpoints run on the free plan. A few, mainly the ones that publish on your behalf, need a paid plan.

Design, 6 operations

delete_designDelete designDELETEPermanently delete a design from the workspace; there is no restore. Files already generated from it are not affected.
get_designGet designGETGet a design: its formats (ids, sizes, preview URLs), every element with the properties it accepts, and its text variables. Read it before generating, to know the layer names and format ids.
get_design_as_importGet design as importGETExport an existing design as the JSON body create_design_import accepts, to clone, version or diff it. Alpha API: the contract may change without notice.
list_designsList designsGETList the designs of the workspace, ordered by name, with optional filters (project, type, orientation, size, format, dates), word search and sort. Without page or per_page every match is returned.
move_designMove designPUTMove a design to another project of the workspace. This resets the design's approval status. List the project ids with list_projects.
rename_designRename designPUTChange the name of one design. The name is trimmed, then must be 1 to 100 characters. Sending the name the design already has succeeds and changes nothing.

Design import, 3 operations

create_design_importCreate design importPOSTCreate a design from a JSON structure (alpha). Give every image, video or audio src as a public URL or data URI: the import is then queued at once. Track it with get_design_import.
get_design_importGet design importGETGet the status of a JSON design import (WAITING_FOR_VALIDATION, QUEUED, PROCESSING, DONE or FAILED), its warnings, its error when FAILED, and the created design in result. Alpha API.
validate_design_importValidate design importPUTQueue an import that is WAITING_FOR_VALIDATION once all its presigned uploads are done; answers 422 missing_assets otherwise. Not needed when every asset was a URL. Takes no body. Alpha API.

Project, 2 operations

create_projectCreate projectPOSTCreate a new project to organize your designs. Assign designs to this project in the Abyssale editor to make them accessible via the API.
list_projectsList projectsGETList the projects of the workspace, ordered by name, optionally filtered by a phrase in the name. Only designs in a project are reachable through the API. Without page or per_page every match is returned.

Workspace template, 2 operations

duplicate_workspace_templateDuplicate workspace templatePOSTDuplicate a workspace template into a project as a new design. Asynchronous: returns duplication_request_id, then call get_duplication_request until status is COMPLETED or ERROR.
list_workspace_templatesList workspace templatesGETList the workspace templates (organisation-level master designs), optionally by category or type. Turn one into a generatable design with duplicate_workspace_template.

Credit, 1 operation

get_creditsGet creditsGETGet the workspace's remaining generation credits and AI credits for the current billing period (available, limit, consumed, extra; available and limit are null on an unlimited plan). Costs nothing.

Design format, 1 operation

get_design_formatGet design formatGETGet one format of a design: size, unit, preview and dynamic image URLs, element layout and variables. Not available on printer_multipage designs.

Dynamic image url, 1 operation

create_dynamic_image_urlCreate dynamic image urlPOSTCreate the dynamic image URL of a static design (one per design; calling again returns the existing one). Variants are then rendered by adding element overrides to that URL's query string.

File, 1 operation

get_fileGet fileGETGet one generated file (image, video, GIF, HTML5 banner or PDF) by id: its download URL (file.url), CDN URL (file.cdn_url), format, design and project.

File export, 1 operation

export_bannersExport bannersPOSTPackage previously generated files (any output type) into one ZIP archive. Returns export_id; the archive URL is delivered only to callback_url (NEW_EXPORT event), no tool reads it back.

Font, 1 operation

list_fontsList fontsGETList the fonts of the workspace (Google Fonts and uploaded fonts), ordered by name, with optional name search and filters. Use a font's id to set the font of a text or button element.

Generation request, 1 operation

get_generation_requestGet generation requestGETGet an asynchronous generation: is_finalized is false while it runs, true once done, with banners[] (file URLs) and errors[] (formats that failed). Kept for 7 days.

Image, 1 operation

generate_imageGenerate imagePOSTGenerate one image from a static design and get its file URL in the response (image.url, file.url, file.cdn_url). Static designs only; one format per call, billed one generation credit.

Media, 1 operation

generate_multi_format_mediaGenerate multi format mediaPOSTQueue generation of several formats of one design (image, video, GIF, HTML5 or PDF). Returns generation_request_id: call get_generation_request with it until is_finalized is true. Each format is billed.

Pdf, 1 operation

generate_multi_page_pdfGenerate multi page pdfPOSTQueue a multi-page print PDF from a printer_multipage design, one entry of pages per page. Returns generation_request_id: call get_generation_request with it until is_finalized is true. Billed by area and page.

Template duplication, 1 operation

get_duplication_requestGet duplication requestGETGet the status of a workspace template duplication started by duplicate_workspace_template: status INIT, IN_PROGRESS, ERROR or COMPLETED, and the new design ids once completed.

Workspace template category, 1 operation

list_workspace_template_categoriesList workspace template categoriesGETThe categories that group your workspace templates. Use an `id` from here as the `category_id` filter on `list_workspace_templates`.

Use Abyssale from Claude Code, Cursor or Codex

Connect Abyssale in LiveContext, create an API key in Settings, then add the LiveContext MCP server to your client. Its Abyssale operations become tools the client can call, next to every other integration you connected.

Claude Code

claude mcp add --transport http livecontext https://livecontext.ai/mcp --header "X-API-Key: YOUR_API_KEY"

Codex (~/.codex/config.toml)

[mcp_servers.livecontext]
url = "https://livecontext.ai/mcp"

[mcp_servers.livecontext.http_headers]
"X-API-Key" = "YOUR_API_KEY"

Cursor, Claude Desktop and other JSON configs

{
  "mcpServers": {
    "livecontext": {
      "type": "http",
      "url": "https://livecontext.ai/mcp",
      "headers": {
        "X-API-Key": "YOUR_API_KEY"
      }
    }
  }
}

Set up the MCP server step by step

How to connect Abyssale

Create your workspace
Sign up free. Nothing to install.
Connect Abyssale once
Add your Abyssale connection in Settings. Every tool from this integration reuses it, and secrets are encrypted at rest and visible only to you.
Describe the job in chat
The workflow builds itself in front of you and picks the endpoints it needs. Drag the steps yourself on the canvas whenever you would rather be explicit.