agentleFS
celeryhqceleryhq/simplified-ai/skills/generate-image/SKILL.md

generate-image

Skill4 starsChanged 4 months agoMITView on GitHub

Create images or broad prompt/reference-guided transformations with Simplified. Use for new photos, graphics, illustrations, posters, and compositions. For background removal/replacement, restoration, upscaling, masked edits, or format conversion, use edit-image.

37 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use scientific-slides, team and 5 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • ZanderRussSkill

    scientific-slides

    Build slide decks and presentations for research talks. Use this for making PowerPoint slides, conference presentations, seminar talks, research presentations, thesis defense slides, or any scientific talk. Provides slide structure, design templates, timing guidance, and visual validation. Works with PowerPoint and LaTeX Beamer.

    Also in 3 projects that use this

  • nicepkgSkill

    team

    根据任务快速组建临时 AI Agent 团队协作。自动从 .claude/agents/ 中选择最合适的成员组队。

    Also in 3 projects that use this

  • AnthropicSkill

    pptx

    Use this skill any time a .pptx or .potx file is involved in any way — as input, output, or both. This includes: creating slide decks, pitch decks, or presentations; reading, parsing, or extracting text from any .pptx or .potx file (even if the extracted content will be used elsewhere, like in an email or summary); editing, modifying, or updating existing presentations; combining or splitting slide files; working with templates (.potx), layouts, speaker notes, or comments. Trigger whenever the user mentions \"deck,\" \"slides,\" \"presentation,\" or references a .pptx or .potx filename, regardless of what they plan to do with the content afterward. If a .pptx or .potx file needs to be opened, created, or touched, use this skill.

    Also in 5 projects that use this

  • AnthropicSkill

    pdf

    Comprehensive PDF manipulation toolkit for extracting text and tables, creating new PDFs, merging/splitting documents, and handling forms. When Claude needs to fill in a PDF form or programmatically process, generate, or analyze PDF documents at scale.

    Also in 4 projects that use this

  • SentrySkill

    skill-creator

    Create new agent skills following the Agent Skills specification. Use when asked to "create a skill", "add a new skill", "write a skill", "make a skill", "build a skill", or scaffold a new skill with SKILL.md. Guides through requirements, writing, registration, and verification.

    Also in 5 projects that use this

  • VercelSkill

    find-skills

    Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skill should be used when the user is looking for functionality that might exist as an installable skill.

    Also in 3 projects that use this

  • +1 more from anthropics

The file

We split this skill into the rules it gives an agent

5 rules. 0 are common, 2 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 31 edited versions we read keep these.

  • 2

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Create or edit an image with Simplified and retain an asset ID when the result will be reused.

    line 10
  2. Very specific:

    Call api_generateImage when the user requests generation.

    line 18
  3. Distinctive:

    Wait for the tool's response, inspect its actual status, and report the result.

    line 19
  4. Very specific:

    Use the normalized public field reference_images for every model.

    line 41
  5. Very specific:

    Read the returned envelope rather than assuming all clients return identical wrappers.

    line 55

“Common” is measured: kept by at least 6 in 10 of the 31 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Generate Image

For background removal/replacement, restoration, upscaling, masked edits, or format conversion, use edit-image. Continue here for prompt/reference-guided generation.

Create or edit an image with Simplified and retain an asset ID when the result will be reused. Use the current generation tool schema as the request contract; model catalogs describe availability and limits, not a reason to substitute provider-specific field names.

Workflow

  1. Resolve workspace/teamspace scope when needed using simplified-workspace. Carry its space_id into related calls.
  2. Resolve existing references with manage-assets: search names/tags with api_listAssets if available, or use supplied IDs/URLs. Check asset readiness with api_getAsset before generating.
  3. Select a compatible model from api_listImageModels, or api_getModelFields(type: "image") if that is the discovery tool exposed in this client. Inspect api_getImageModelFields or the available model-fields tool for limits/options when the request requires a particular resolution, quality, reference count, or provider. Honor explicit model choices when supported. Model IDs and credit costs can change; do not guess them.
  4. Build the nested request below. Use storage: "asset" for reusable outputs; use transient only for an explicitly temporary result. The current schema recommends/defaults to asset storage; pass the choice explicitly.
  5. Call api_generateImage when the user requests generation. Generation spends credits: clarify once when cost, quantity, or intent is ambiguous. Do not add a confirmation when the user already authorized that generation.
  6. Wait for the tool's response, inspect its actual status, and report the result. Current apikit waits for completion; do not start a second job or manually poll while the call is still active.

Current request contract

{
  "model": "<model ID returned by discovery>",
  "storage": "asset",
  "parameters": {
    "prompt": "A studio product photo with space for a headline",
    "aspect_ratio": "4:5",
    "count": 1
  }
}

This is a template: substitute a discovered model ID before calling it. model, storage, and parameters are top-level. prompt, aspect_ratio, count, and reference_images belong inside parameters.

For editing, add parameters.reference_images: ["<ready asset UUID>"]. For composition, supply several references in that array; supported counts depend on the model. Prefer workspace asset UUIDs. The image tool also accepts absolute HTTP(S) image URLs; preserve the full signed query string if using one. Local paths, filenames, and another workspace's UUIDs are invalid.

capability is optional: no references selects prompt, one selects reference_image, and several select multiple_images. Several references compose one output; use count for several generated outputs. If setting a capability explicitly, confirm the model supports it.

Use the normalized public field reference_images for every model. Do not replace it with provider fields such as input_image or source_image, even if a catalog describes internal provider fields. Add optional model-specific controls only when the generation schema/live contract accepts them.

Choose for the outcome

For ordinary social graphics, select a currently available general model; for dense text/layout or higher resolution, inspect compatible models and limits. For continuity/editing, filter by reference support. When the user specifies a provider, use that family's currently recommended compatible ID. For a budget-sensitive request, inspect available credit metadata and generate only the requested count. Treat credit rates as estimates unless the tool explicitly guarantees the charge; size, quality, and count can affect usage.

Storage and results

ModeUse
assetSave in the Asset Library and retain a reusable UUID
transientTemporary output URL without a permanent asset ID
defaultSave in the image gallery; do not assume this supplies an Asset Library UUID

Read the returned envelope rather than assuming all clients return identical wrappers. Common successful image results are in detail.result; entries can be URL strings (transient) or objects containing asset_id and url (asset). Keep every returned output when generating multiple images.

Present the result using the client’s supported preview or a usable link and include the permanent asset_id when present. Signed file URLs may expire even for saved assets. Refresh them with api_getAsset. If the user later wants to reuse a transient image, import its accessible URL with api_createAsset instead of generating again.

For social drafts, hand off actual IDs to simplified-social.media; generation is not permission to publish.

Errors and incomplete jobs

An accepted request is not proof of a finished image. On timeout/pending status, retain the returned job identifiers and use only the continuation operation the response/tool schema specifies. Never blindly repeat a paid generation. Report provider errors and use the supported-value list for a targeted correction. A 429 may indicate quota or rate limits: report its actual message rather than assuming credits are always exhausted.

Examples users can ask: “Make a vertical product photo and save it”; “Find my uploaded logo and use it in a poster”; “Edit this asset to use a blue background”; “Create three variations for review.”

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.