跳到正文
cccskills
免费在 GitHub 公开

canva

Create and edit Canva designs for presentations, campaigns, flyers, and social posts. Generate standalone images, remove backgrounds, recover editable layers from flat graphics, and use available brand templates or resize existing designs.

中文简介正在准备中。当前显示英语简介。

原文查看原文说明

Create and edit Canva designs, generate images, remove backgrounds, recover editable layers, and work with Canva libraries. Use for branded presentations and slide decks, campaign artwork, flyers and banners, applying saved brand kits and templates, and resizing designs for social posts and stories through Canva's official MCP server.

集成工具Canva
查看安装方法

适用场景

  • Create a flyer or presentation in Canva
  • Make campaign artwork from a brand template
  • Resize a design for social posts and stories
  • Remove backgrounds or separate image layers

包含的文件(3)

  • SKILL.md8.2 KB
  • connector_auth_config.yaml220 B
  • manifest.yaml3.4 KB

SKILL.md(原文)

安装前可以查看交给 AI 代理的指令内容。

Canva

Use the installed canva CLI. Start with canva status. If it reports not_connected, run canva authorize-url and share only the returned connect_url. If the CLI reports outdated OAuth settings, have the user disconnect Canva in Settings and reconnect. Never request tokens in chat.

Run canva list-tools for live schemas and use only tools returned there. Follow their input schemas exactly and include a concise user_intent. hatch_permission_overrides identifies argument-dependent permissions; opening an editing transaction is a write and deleting pages is a delete. Some tools require Canva Pro, Enterprise, or available AI credits.

Save each raw response before parsing. Check result.isError, then read result.structuredContent when present or parse the JSON text block in result.content. Both response forms are valid. A local parsing failure does not mean a mutation failed: recover its response or inspect saved state before continuing. Never repeat a copy, creation, or commit just to obtain its output.

canva search-designs [--query <keywords>] [--continuation <token>]
canva get-design-content --design-id <id>
canva get-design-pages --design-id <id>
canva list-tools
canva call-tool --name <tool-name> --arguments-json '<JSON object>'

Create designs and images

Use create-design for a new editable layout: social post, infographic, poster, flyer, presentation, document, or sheet. Put all necessary source facts and requested wording in brief; the tool does not inherit chat or connected-source context. Fetch the user-selected source first. For an exact size, state dimensions and orientation in the brief. If supplying format, include the orientation; an ambiguous format may resolve to a square. Pass outline only for a presentation outline the user supplied or approved.

Use generate-image for an explicitly requested standalone image, photo, illustration, artwork, or image of an infographic. Use one MEDIA reference per uploaded source image. This generates an image rather than an editable page layout. Use its returned media_id when editing a generated image again.

create-design cannot apply brand kits or brand templates. For an explicitly on-brand request, use the available legacy generate-design, get-design-candidates, and create-design-from-candidate flow, showing candidates for the user's choice. These tools are otherwise deprecated when create-design is listed; failure of create-design is not permission to fall back. generate-design does not support presentations despite its enum. The legacy outline-review widget and structured presentation generator are not exposed by this CLI. Do not substitute a non-branded generation when a brand-specific presentation workflow is unavailable. A selected brand template can still be copied with create-design-from-brand-template or autofilled.

Creation and layer separation return asynchronous jobs. When no Canva widget is shown (including CLI use), poll the corresponding get-create-design-async-job, get-generate-image-job, or get-separate-image-layers-job. Honor every returned wait interval and updated continuation token. Never start a replacement write because a job is pending. Stop on terminal failure; respect quota and moderation failures. Show completed results using the preview workflow below. For generated images, include the returned Canva upload link with the text Open generated image.

Upload and transform images

For attachments, local files, or generated files up to 256 MiB, use canva upload-file --file <path> --user-intent '<purpose>'. Its approval shows the selected file and an image preview for workspace images. The CLI obtains a single-use upload URL and sends the raw bytes after approval; use the returned resource IDs in later calls. Do not repeat an upload whose outcome is unknown. The low-level create-upload-url flow remains available for larger files: one raw-byte POST with Content-Type: application/octet-stream, no multipart, JSON, or base64. Never retry a consumed upload URL.

upload-asset-from-url and import-design-from-url accept already-public HTTPS sources only. Do not publish local or private files to use those tools. Uploading media does not place it into a design. create-design has no asset input parameter: when exact supplied media must appear, use the editing flow to insert or replace media with its verified Canva ID, then inspect the result.

Use remove-background with an already-uploaded MEDIA reference to produce a new image with transparent alpha. It does not replace the scene or crop the subject. Use separate-image-layers with an uploaded image's asset_id to turn a flat graphic into a new editable design; the original stays unchanged. Verify editable elements with read-design, not appearance alone.

Read, edit, and organize

Use read-design for metadata, text, page metadata, thumbnails, and presenter notes. The old MCP names get-design, get-design-content, get-presenter-notes, and get-design-thumbnail are removed. The CLI get-design-content convenience command now calls read-design. get-design-pages remains available for saved page previews.

To edit, call read-design with open_transaction: true and include thumbnails in filter.fields for a before preview. Use its transaction ID, element locators, and page flags in edit-design with finalize: keep_open. Read that transaction to inspect unsaved changes. Show the preview and obtain explicit approval before edit-design with finalize: commit and no operations. Use finalize: cancel to discard edits. These replace start-editing-transaction, perform-editing-operations, commit-editing-transaction, and cancel-editing-transaction, even when older server descriptions still mention those names. Cancel stale transactions and open a fresh one.

merge-designs combines or reorders whole pages. Obtain explicit approval of the exact operations before each call; deleting pages is permanent. copy-design and resize-design create new designs and preserve their source. Before autofill-design, inspect get-design-dataset or get-brand-template-dataset and match its field names/types. Set update_in_place only when the user requested overwriting that design.

For brand-template updates, start with create-brand-template-draft, edit and save its design through the current transaction flow, then use publish-brand-template only when the user requested organization-wide publication. Publishing affects a reusable shared template.

Resolve Canva shortlinks before using designs. Confirm get-export-formats before export-design. Use help for current Canva product support questions, not to describe this CLI's capabilities.

Show results

Show a visual preview in chat alongside the returned Canva link when delivering designs or images. Save returned image content, or download a returned thumbnail URL with curl, to a file under workspace/. Inspect it with read, then attach it on its own line as ![Preview](sandbox://workspace/path/to/image.png). Preview each design in a small set; for a long deck, show representative pages and label their page numbers. Local attachments remain useful after signed preview URLs expire.

Check that the image shows the expected content; HTTP success alone does not rule out a blank or stale thumbnail. If needed, request a fresh thumbnail or export the saved design in a supported image format. Never commit unsaved edits just to obtain a preview, or present a saved export as an unsaved draft. If a usable preview is unavailable, explain that and keep the Canva link; do not repeat creation or editing to repair a preview.

When asked to list designs, omit --query and follow continuation tokens. This does not authorize background crawling or bulk indexing. Comments and replies are visible to collaborators; post only when the user requested them.

评价

还没有评价。欢迎分享你的使用感受。

评论

无需登录即可发表。欢迎分享使用感受、提问,或说明在哪些环境下无法运行。

还没有评论。来写第一条吧。

提交后会立即公开。请勿填写邮箱、电话等个人信息。URL 最多 2 个。

0 / 1000 字

此表单受 reCAPTCHA 保护,适用 Google 的隐私权政策和服务条款。

同一仓库的技能

3d-object

免费

Create a 3D object with three.js, inspect it from every angle in an interactive viewer, and download it as OBJ + MTL or GLB for use in other 3D applications.

中文简介正在准备中。当前显示英语简介。

  • Create a 3D object and inspect it from every angle
  • Export a model for further editing in Blender
  • Review an object's silhouette and materials
asgeirtj/system_prompts_leaks7万2026年10月20日 更新

Maintain a private task list with deadlines, next steps, waiting parties, and blockers. Update progress and dependencies, merge duplicate items, and archive closed tasks so ongoing commitments stay clear and easy to follow.

中文简介正在准备中。当前显示英语简介。

  • Record tasks and commitments with deadlines
  • Track pending replies and reasons work cannot proceed
  • Update task dependencies when prerequisites are completed
asgeirtj/system_prompts_leaks7万2026年10月20日 更新

agents

免费

Coordinate a task from a verifiable goal through completion. Handle work directly or split independent tasks across separate agent sessions, track progress, verify results, and arrange pull request merging when included in the goal.

中文简介正在准备中。当前显示英语简介。

  • Split independent tasks across agents
  • Check progress across ongoing tasks
  • Resume a previously started project
asgeirtj/system_prompts_leaks7万2026年10月20日 更新

Create animated HTML pages with smooth scene transitions and playback controls. An editable timeline lets users adjust scene lengths and speeds, with fixed landscape or portrait layouts prepared for video export.

中文简介正在准备中。当前显示英语简介。

  • Create an animated product walkthrough
  • Explain a message through moving text and graphics
  • Build a portrait animation for video export
asgeirtj/system_prompts_leaks7万2026年10月20日 更新

Review synced Apple Health records, including steps, heart rate, sleep, and workouts. Retrieve daily summaries or detailed samples for a chosen period, and check sync status and missing data before interpreting the results.

中文简介正在准备中。当前显示英语简介。

  • Review daily steps and heart rate over time
  • Check sleep stages, duration, and awakenings
  • Review workout duration, distance, and calories
asgeirtj/system_prompts_leaks7万2026年10月20日 更新

Create and edit Word documents and templates, reorganize text, replace images, and work with tracked changes or comments. Apply proper headings and lists, then render every page to check its appearance before delivery.

中文简介正在准备中。当前显示英语简介。

  • Create a structured Word document
  • Reorganize content in an existing document
  • Replace images or text in a document
asgeirtj/system_prompts_leaks7万2026年10月20日 更新

查看 asgeirtj 的全部技能

报告此技能的问题