본문으로 건너뛰기
cccskills
무료GitHub에 공개

canva-automation

Automate Canva tasks via Rube MCP (Composio): designs, exports, folders, brand templates, autofill. Always search tools first for current schemas.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

연동 도구Canva
설치 방법 보기

포함된 파일(1)

  • SKILL.md8.1 KB

SKILL.md(원문)

설치하기 전에 에이전트에게 전달되는 지시 내용을 확인할 수 있습니다.

Canva Automation via Rube MCP

Automate Canva design operations through Composio's Canva toolkit via Rube MCP.

Toolkit docs: composio.dev/toolkits/canva

Prerequisites

  • Rube MCP must be connected (RUBE_SEARCH_TOOLS available)
  • Active Canva connection via RUBE_MANAGE_CONNECTIONS with toolkit canva
  • Always call RUBE_SEARCH_TOOLS first to get current tool schemas

Setup

Get Rube MCP: Add https://rube.app/mcp as an MCP server in your client configuration. No API keys needed — just add the endpoint and it works.

  1. Verify Rube MCP is available by confirming RUBE_SEARCH_TOOLS responds
  2. Call RUBE_MANAGE_CONNECTIONS with toolkit canva
  3. If connection is not ACTIVE, follow the returned auth link to complete Canva OAuth
  4. Confirm connection status shows ACTIVE before running any workflows

Core Workflows

1. List and Browse Designs

When to use: User wants to find existing designs or browse their Canva library

Tool sequence:

  1. CANVA_LIST_USER_DESIGNS - List all designs with optional filters [Required]

Key parameters:

  • query: Search term to filter designs by name
  • continuation: Pagination token from previous response
  • ownership: Filter by 'owned', 'shared', or 'any'
  • sort_by: Sort field (e.g., 'modified_at', 'title')

Pitfalls:

  • Results are paginated; follow continuation token until absent
  • Deleted designs may still appear briefly; check design status
  • Search is substring-based, not fuzzy matching

2. Create and Design

When to use: User wants to create a new Canva design from scratch or from a template

Tool sequence:

  1. CANVA_ACCESS_USER_SPECIFIC_BRAND_TEMPLATES_LIST - Browse available brand templates [Optional]
  2. CANVA_CREATE_CANVA_DESIGN_WITH_OPTIONAL_ASSET - Create a new design [Required]

Key parameters:

  • design_type: Type of design (e.g., 'Presentation', 'Poster', 'SocialMedia')
  • title: Name for the new design
  • asset_id: Optional asset to include in the design
  • width / height: Custom dimensions in pixels

Pitfalls:

  • Design type must match Canva's predefined types exactly
  • Custom dimensions have minimum and maximum limits
  • Asset must be uploaded first via CANVA_CREATE_ASSET_UPLOAD_JOB before referencing

3. Upload Assets

When to use: User wants to upload images or files to Canva for use in designs

Tool sequence:

  1. CANVA_CREATE_ASSET_UPLOAD_JOB - Initiate the asset upload [Required]
  2. CANVA_FETCH_ASSET_UPLOAD_JOB_STATUS - Poll until upload completes [Required]

Key parameters:

  • name: Display name for the asset
  • url: Public URL of the file to upload (for URL-based uploads)
  • job_id: Upload job ID returned from step 1 (for status polling)

Pitfalls:

  • Upload is asynchronous; you MUST poll the job status until it completes
  • Supported formats include PNG, JPG, SVG, MP4, GIF
  • File size limits apply; large files may take longer to process
  • The job_id from CREATE returns the ID needed for status polling
  • Status values: 'in_progress', 'success', 'failed'

4. Export Designs

When to use: User wants to download or export a Canva design as PDF, PNG, or other format

Tool sequence:

  1. CANVA_LIST_USER_DESIGNS - Find the design to export [Prerequisite]
  2. CANVA_CREATE_CANVA_DESIGN_EXPORT_JOB - Start the export process [Required]
  3. CANVA_GET_DESIGN_EXPORT_JOB_RESULT - Poll until export completes and get download URL [Required]

Key parameters:

  • design_id: ID of the design to export
  • format: Export format ('pdf', 'png', 'jpg', 'svg', 'mp4', 'gif', 'pptx')
  • pages: Specific page numbers to export (array)
  • quality: Export quality ('regular', 'high')
  • job_id: Export job ID for polling status

Pitfalls:

  • Export is asynchronous; you MUST poll the job result until it completes
  • Download URLs from completed exports expire after a limited time
  • Large designs with many pages take longer to export
  • Not all formats support all design types (e.g., MP4 only for animations)
  • Poll interval: wait 2-3 seconds between status checks

5. Organize with Folders

When to use: User wants to create folders or organize designs into folders

Tool sequence:

  1. CANVA_POST_FOLDERS - Create a new folder [Required]
  2. CANVA_MOVE_ITEM_TO_SPECIFIED_FOLDER - Move designs into folders [Optional]

Key parameters:

  • name: Folder name
  • parent_folder_id: Parent folder for nested organization
  • item_id: ID of the design or asset to move
  • folder_id: Target folder ID

Pitfalls:

  • Folder names must be unique within the same parent folder
  • Moving items between folders updates their location immediately
  • Root-level folders have no parent_folder_id

6. Autofill from Brand Templates

When to use: User wants to generate designs by filling brand template placeholders with data

Tool sequence:

  1. CANVA_ACCESS_USER_SPECIFIC_BRAND_TEMPLATES_LIST - List available brand templates [Required]
  2. CANVA_INITIATE_CANVA_DESIGN_AUTOFILL_JOB - Start autofill with data [Required]

Key parameters:

  • brand_template_id: ID of the brand template to use
  • title: Title for the generated design
  • data: Key-value mapping of placeholder names to replacement values

Pitfalls:

  • Template placeholders must match exactly (case-sensitive)
  • Autofill is asynchronous; poll for completion
  • Only brand templates support autofill, not regular designs
  • Data values must match the expected type for each placeholder (text, image URL)

Common Patterns

Async Job Pattern

Many Canva operations are asynchronous:

1. Initiate job (upload, export, autofill) -> get job_id
2. Poll status endpoint with job_id every 2-3 seconds
3. Check for 'success' or 'failed' status
4. On success, extract result (asset_id, download_url, design_id)

ID Resolution

Design name -> Design ID:

1. Call CANVA_LIST_USER_DESIGNS with query=design_name
2. Find matching design in results
3. Extract id field

Brand template name -> Template ID:

1. Call CANVA_ACCESS_USER_SPECIFIC_BRAND_TEMPLATES_LIST
2. Find template by name
3. Extract brand_template_id

Pagination

  • Check response for continuation token
  • Pass token in next request's continuation parameter
  • Continue until continuation is absent or empty

Known Pitfalls

Async Operations:

  • Uploads, exports, and autofills are all asynchronous
  • Always poll job status; do not assume immediate completion
  • Download URLs from exports expire; use them promptly

Asset Management:

  • Assets must be uploaded before they can be used in designs
  • Upload job must reach 'success' status before the asset_id is valid
  • Supported formats vary; check Canva documentation for current limits

Rate Limits:

  • Canva API has rate limits per endpoint
  • Implement exponential backoff for bulk operations
  • Batch operations where possible to reduce API calls

Response Parsing:

  • Response data may be nested under data key
  • Job status responses include different fields based on completion state
  • Parse defensively with fallbacks for optional fields

Quick Reference

TaskTool SlugKey Params
List designsCANVA_LIST_USER_DESIGNSquery, continuation
Create designCANVA_CREATE_CANVA_DESIGN_WITH_OPTIONAL_ASSETdesign_type, title
Upload assetCANVA_CREATE_ASSET_UPLOAD_JOBname, url
Check uploadCANVA_FETCH_ASSET_UPLOAD_JOB_STATUSjob_id
Export designCANVA_CREATE_CANVA_DESIGN_EXPORT_JOBdesign_id, format
Get exportCANVA_GET_DESIGN_EXPORT_JOB_RESULTjob_id
Create folderCANVA_POST_FOLDERSname, parent_folder_id
Move to folderCANVA_MOVE_ITEM_TO_SPECIFIED_FOLDERitem_id, folder_id
List templatesCANVA_ACCESS_USER_SPECIFIC_BRAND_TEMPLATES_LIST(none)
Autofill templateCANVA_INITIATE_CANVA_DESIGN_AUTOFILL_JOBbrand_template_id, data

Powered by Composio

리뷰

아직 리뷰가 없습니다. 사용해 본 소감을 남겨 주세요.

댓글

로그인 없이 쓸 수 있습니다. 사용 후기, 질문, 작동하지 않은 환경 등을 편하게 남겨 주세요.

아직 댓글이 없습니다. 첫 댓글을 남겨 보세요.

보내면 바로 공개됩니다. 이메일 주소나 전화번호 같은 개인 정보는 쓰지 마세요. URL은 2개까지입니다.

0 / 1,000자

이 양식은 reCAPTCHA로 보호되며 Google 개인정보처리방침 및 서비스 약관이 적용됩니다.

같은 저장소의 스킬

Automate ActiveCampaign tasks via Rube MCP (Composio): manage contacts, tags, list subscriptions, automation enrollment, and tasks. Always search tools first for current schemas.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

davepoon/buildwithclaude3.7천2026년 10월 19일 업데이트

Pull the ads a competitor is running right now on Meta (Facebook/Instagram) and Google, then extract the hooks, offers, CTAs and landing pages worth modelling. Use when writing ad copy, planning creative tests, choosing a landing page angle, or when the user asks "what is X advertising", "what ads does X run", "find competitor ad hooks". Reads live Ad Library and Ads Transparency data through the Tracetify MCP; falls back to a manual Ad Library walkthrough when the connector is absent.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

davepoon/buildwithclaude3.7천2026년 10월 19일 업데이트

Analytics your AI agent can actually use. Track, analyze, run A/B experiments, and optimize across all your projects via CLI. Includes a growth playbook so your agent knows HOW to grow, not just what to track.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

davepoon/buildwithclaude3.7천2026년 10월 19일 업데이트

Teaches when to recall from long-term memory before acting and when to save durable decisions, corrections and failures afterwards. Use when a memory tool or MCP memory server is connected but the agent is not using it consistently, when the user complains that the assistant forgets preferences, conventions or past decisions between sessions, or when setting up persistent memory for a project. Works with any memory backend: a folder of Markdown files, a local MCP server, or a managed service.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

davepoon/buildwithclaude3.7천2026년 10월 19일 업데이트

Audit local AI coding-agent sessions with agenttrace for cost, tokens, tool failures, latency, anomalies, health, diffs, and CI gates.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

davepoon/buildwithclaude3.7천2026년 10월 19일 업데이트

Checks which private files (.env, keys, secrets) Claude Code and its subagents reached in this project, whether a value came into an agent's context, and which key to change, by running the open source agentwhy CLI over the session records Claude Code already keeps on disk. Use when the user asks whether the agent or a subagent read their .env or other secrets, which private files a session touched, whether a key leaked into the conversation, or which key to rotate. Runs locally and never prints a secret value.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

davepoon/buildwithclaude3.7천2026년 10월 19일 업데이트

davepoon의 스킬 모두 보기

이 스킬의 문제 신고