Image Generation
Route text-to-image, edits, and batch variant requests through task classification and model policy before PostPlus executes jobs.
Overview
Image Generation is an agent skill most often used in Build (also Launch content, Grow content) that classifies image requests and hands them to image-batch-runner without submitting jobs itself.
Install
npx skills add https://github.com/postplusai/postplus-skills --skill image-generationWhat is this skill?
- Classifies image work into text_to_image, image_edit, reference-driven, product, persona, banner, thumbnail, and storybo
- Enforces a hard boundary: identifies inputs and policies but does not submit generation jobs
- Hands off normalized requests to image-batch-runner after model and reference rules are chosen
- Covers batch variants, product URLs, research handoffs, and uploaded reference images
- Pairs with reference-decode and reference-contract-builder when boundaries need decoding first
- Four-step controller flow: input type, task class, model/reference rules, runner handoff
- Explicit task-class table covering text_to_image, edits, and reference-driven generation paths
Adoption & trust: 1 installs on skills.sh; 8 GitHub stars; 3/3 security scanners passed (skills.sh audits); trending (+100% hot-view momentum).
What problem does it solve?
You have a messy image or batch request and no safe way to pick task class, references, and model rules before your agent hits the generator.
Who is it for?
Indie builders wiring PostPlus image pipelines who need a single gate before batch execution.
Skip if: Sessions where jobs are already normalized, the final asset is video or audio, or you only need hook decoding without generation routing.
When should I use this skill?
User wants text-to-image, image edit, reference-image generation, product or persona images, banners, thumbnails, storyboards, or batch variants and the next step is classification—not execution.
What do I get? / Deliverables
You get a normalized controller handoff with task class and reference policy so image-batch-runner can execute the batch.
- Classified image task with model and reference policy
- Structured handoff payload for image-batch-runner
Recommended Skills
Journey fit
Spans multiple journey phases - primary shelf plus alternate fits below.
Image workflows are authored and wired into agent stacks during product build, even when assets ship later for launch or growth. Controller skills that normalize prompts, references, and handoffs belong on the agent-tooling shelf alongside runners and contract skills.
Where it fits
Wire a Claude Code skill chain so persona and product image requests never call the API until task class and references are set.
Normalize banner and thumbnail specs from a launch brief before batching variants for social and landing pages.
Turn recurring content themes into classified image batches with consistent reference rules for a newsletter or blog cadence.
How it compares
Use as the pre-execution controller, not as a replacement for image-batch-runner or reference-contract-builder.
Common Questions / FAQ
Who is image-generation for?
Solo and indie builders using PostPlus skills in Claude Code, Cursor, or Codex who generate marketing, product, or creative images through an agent and want classification before any API spend.
When should I use image-generation?
Use it in Build when wiring agent-tooling for creative pipelines; at Launch when producing banners and thumbnails; and in Grow when batching content variants—always before image-batch-runner runs.
Is image-generation safe to install?
It is documentation-style routing logic with no job submission in this skill; review the Security Audits panel on this Prism page before enabling it in a repo with secrets or production keys.
Workflow Chain
Requires first: reference decode, reference contract builder
Then invoke: image batch runner
SKILL.md
READMESKILL.md - Image Generation
# Image Generation ## Use When - The desired final asset is an image or image batch. - The request may include text prompts, uploaded images, previous outputs, product URLs, research handoffs, product images, persona images, banners, thumbnails, storyboard panels, or batch variants. - The next decision is task class, model/reference policy, and runner handoff. ## Do Not Use When - The final asset is video, audio, subtitles, transcripts, or an edit plan. - The image request is already normalized and ready to execute. Use `image-batch-runner`. - The user needs hook decoding or reference boundaries first. Use `reference-decode` or `reference-contract-builder`. ## Core Boundary This is the image generation controller. It does not submit jobs. It must: 1. identify the input type, 2. classify the image task, 3. select model and reference rules, 4. create the handoff for `image-batch-runner`. ## Task Classes | Task class | Typical input | Handoff | | --- | --- | --- | | `text_to_image` | prompt only | write normalized image brief for `image-batch-runner` | | `image_edit` | uploaded image plus change request | bind edit image and preserve/alter rules | | `reference_image` | benchmark frame or style board | use `reference-contract-builder` before runner | | `product_image` | product photo, URL, or product facts | bind product identity and forbid invented claims | | `banner_thumbnail` | offer, hook, platform | require aspect ratio and text/UI policy | | `storyboard_image` | panel plan or board spec | hand storyboard panels to `image-batch-runner` | | `batch_variant` | many variants or personas | preserve shared rules and vary only declared fields | ## Model And Reference Rules - Text-only drafts may use a text-to-image endpoint. - Edits require a bound source image and an edit-capable endpoint. - Persona, product, and brand identity are `binding` references unless the user explicitly marks them inspiration-only. - Benchmark clips, mood boards, and competitor references are inspiration-only unless the contract says otherwise. - Excluded references must not enter the runner request. ## Routing Table | If not image-generation | Send to | | --- | --- | | Needs media understanding first | `media-router` | | Needs reference meaning decoded | `reference-decode` | | Needs binding/inspiration/excluded contract | `reference-contract-builder` | | Needs storyboard panels | `storyboard-grid-writer` | | Needs execution with normalized request | `image-batch-runner` | | Final output is video | `video-generation` | | Final output is audio | `audio-generation` | ## Output Shape Return: - `taskClass` - `inputType` - `modelRule` - `referencePolicy` - `requiredRunnerInputs` - `handoffSkill` - `mustNotDo` ## Stop Conditions - Stop when required user intent, source evidence, or owned input artifacts are missing and guessing would change the result. - Do not let `image-batch-runner` make creative classification decisions. - If an owned CLI or script command fails, report the exact error and stop. Do not bypass the failure with metadata-only answers, readiness probing, local payload rewrites, fallback providers, or unpublished tools. ## Public Command Boundary - Choose the smallest matching command or workflow from the user input and run it directly. - If an owned CLI or script command fails, report the exact error and stop. Do not bypass the failure with metadata-only answers, readiness probing, local payload rewrites, fallback providers, or unpu