- Home
- Skills
- Image & Video
- CoverCraft Pro — Editable Thumbnails & Covers
More screenshots
Works with the AI Agents you already use
CoverCraft Pro — Editable Thumbnails & Covers
Create distinctive thumbnails and covers with precise bilingual text, editable layouts and format recomposition.
$2.99
CoverCraft Pro — Editable Thumbnails & Covers
Example session with this skill installed
Create an English tutorial thumbnail for a lesson about editing a cover and exporting PNG/JPEG. Use the actual editor screenshot and the fictional unbranded headphone asset as separate images. Use the exact hook: EDIT. / EXPORT. / DONE. Deliver a 1280 x 720 landscape PNG. Preserve the artwork-preview area in the screenshot and make the hook readable at 160 pixels.
- Read your context and instructions
- Compiled the covercraft pro
- Generated the image
Finished cover: EDIT. EXPORT. DONE.
Creator-side demonstration using a fictional brief and the actual locally running editor. This is not a verified Agensi buyer run.

The artwork
A 1280 x 720 landscape cover uses a cool navy field, a bold three-line ivory headline, a framed screenshot of the real editor, and a separately positioned fictional headphone image. Supporting copy reads: A real editable cover project. / PNG and JPEG export. The footer identifies this as a local editor screenshot.
Exact copy
- Hook: EDIT. / EXPORT. / DONE.
- Eyebrow: WORKFLOW / ACTUAL SOFTWARE.
- Support: A real editable cover project. / PNG and JPEG export.
- Footer: COVERCRAFT / LOCAL EDITOR SCREENSHOT.
Production decisions
The action headline carries the main message at small size; the screenshot supplies visual proof rather than requiring the viewer to read tiny interface text. The screenshot and product are separate image layers. A manually marked source-image rectangle protects the actual artwork-preview portion of the screenshot. The fictional product does not imply a real brand or endorsement.
Actual delivered files
- main.png: exported PNG, 1280 x 720.
- main.jpg: exported JPEG, 1280 x 720.
- main.html: locally generated editable artwork with two embedded images.
- main.json: creator-side input brief and independently selected image IDs.
Observed acceptance
The actual PNG was inspected at full size and 160 pixels wide. Both raster files decode at the requested dimensions. Detected text bounds, text collisions and annotated crop/overlap checks pass. Editing the hook to EDIT. / SAVE. / DONE., saving the artwork and reopening it retained both image assets and the protection rectangle; the reopened PNG matches the pre-save pixels.
Generated files above describe the creator-side run. Buyer delivery of editable HTML and optional helper resources remains subject to Agensi permission and an actual MCP acceptance run. No CTR lift or buyer-completion claim is made.
1791885088540-1nrx.png
PNG · 1536×1024
Example file from a real run - the skill writes it into your workspace.
Connects securely to your tools. The creator never sees your data.
About this skill
Turn a topic, portrait, product image or reference into a finished cover with a clear visual idea. CoverCraft Pro guides your Agent through hook selection, art direction, original imagery, precise typography, composition, inspection and revision.
Current release: 1.1.0.
What you can create
- Video thumbnails, social/content covers, course artwork and creator brand series.
- English, Chinese and intentionally bilingual artwork, with exact text composed separately from imagery.
- Distinct creative directions that change composition, typography, imagery and material.
- Landscape, portrait, vertical, square and custom formats, with a newly arranged hierarchy for each requested ratio.
- A/B creative hypotheses with different hooks; use your own platform analytics to measure results.
Production depth
Sixteen starting art directions are available in the optional local helper, alongside bespoke composition. The helper supports up to eight separately positioned raster images, editable custom layers, source-image protection annotations, 160/320-pixel previews and actual PNG/JPEG export. Its protection check uses manually marked rectangles, not automatic face recognition. Saving and reopening generated editable artwork is supported locally, subject to the delivery channel's permission.
Eight fictional scene studies demonstrate creator tutorials, actual software proof, courses, comparisons, an original game world, reading/editorial work, bilingual events and a consistent brand series. Fourteen cover outputs have been exported and visually reviewed. A separately authored reading composition demonstrates a workflow without the optional builder.
Requirements
Your Agent supplies the model and tools: image generation/editing or authorized source images, plus browser/file capabilities for artwork inspection and export. The optional self-contained helper uses Python's standard library and a modern browser; it needs no npm, external JavaScript, seller backend or bundled API key. Image-generation credits and third-party subscriptions are not included.
Delivery and verification
This is a one-time Skill licence delivered through Agensi MCP. Skill files and helper source are not offered as buyer downloads. The latest approved instructions load through your connected Agent; use depends on Agensi service availability.
Creator-side acceptance verified six Python tests, 160 recipe/ratio/language geometry combinations, real PNG/JPEG exports, multi-image editing, and matching pixels after saved-artwork reopening. Actual Agensi buyer execution, retrieval of optional files/binary assets, and permission to save generated editable HTML have not yet been verified. Platform review is requested for these points. Helper-dependent capabilities apply only where the platform supplies and permits their execution.
No click-through uplift, real-person likeness benchmark or customer success rate is claimed.
Agensi MCP buyer execution and optional-file/binary delivery are not yet verified. Generated editable HTML may be saved only if permitted by the platform. Photo work depends on your Agent image tools or supplied assets; no credits are included. Face/subject protection is manually annotated and checks text bounds and rectangular crop loss; shape/image occlusion needs visual inspection. System fonts can differ across devices. A/B candidates are creative hypotheses, not measured CTR results. No real-person likeness benchmark has been conducted.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 1 month ago
Permissions
File Scopes
Read only user-authorized input assets; write only the chosen generated-artwork directory. covercraft-output/** is an illustrative output scope. Never save the loaded Skill or protected helper source. Browser is used for inspection/export; shell/Python is optional when the platform supplies and permits helper execution. Network is optional for user-selected image providers and public reference research; no fixed seller endpoint, API key or mandatory account is required. Agensi MCP optional-file delivery and generated editable-HTML persistence remain unverified.