Skip to main content
skill CX Codex

oma-image · first-fluke — Claude Code skill: what it does, how to install

oma-image routes image generation requests across Codex, Pollinations and Gemini, and turns a reference picture into a new variation matching its style or layout.

MIT ★ 1,472 updated 2026-11-12

Install into Claude Code · Plugin marketplace All methods and platforms ↓
/plugin marketplace add first-fluke/oh-my-agent
/plugin install oma@oh-my-agent
Open on GitHub

What oma-image does

The skill wraps oma image generate and dispatches the prompt to whichever provider CLI is authenticated: Codex, Pollinations or Gemini. Supply a reference image and, if the chosen vendor supports it, the path gets passed through automatically to produce a variation that keeps the subject, lighting or composition of the original. Passing --vendor all runs every vendor for side-by-side comparison, but every one of them has to be available first.

Each run drops a manifest.json next to the output files with the prompt, vendor, model and other details needed to reproduce the result later. Runs estimated at $0.20 or more ask for confirmation before executing; in CI or other non-interactive setups without a TTY, the command just exits with code 1 and points to the --yes flag instead of prompting. A single call tops out at five images, and writing outside $PWD needs --allow-external-output.

Pixel editing, masking, inpainting, cropping or format conversion fall outside its job: the skill flags those requests and points toward an image editor instead.

What it helps with

  • generate an image through Codex, Pollinations or Gemini from one command
  • produce a reference-guided variation that keeps style and composition
  • run the same prompt across multiple vendors for comparison
  • get a manifest.json with parameters to reproduce a generation
  • check projected cost with --dry-run before running a batch

Example prompts for the agent

  • Generate a product shot on a white background with oma-image
  • Create a variation of this reference keeping the same lighting and style
  • Run this prompt across all vendors and compare the results

How to install oma-image

1Plugin marketplace

/plugin marketplace add first-fluke/oh-my-agent
/plugin install oma@oh-my-agent

This installs the whole oma plugin with all its skills

2Skills CLI (Vercel)

npx skills add first-fluke/oh-my-agent --skill oma-image -a claude-code

3Manually: put the files in place

~/.claude/skills/oma-image/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/oma-image/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about oma-image

How do I install oma-image in Claude Code?

Through the marketplace: run /plugin marketplace add first-fluke/oh-my-agent, then /plugin install oma@oh-my-agent.

What does oma-image do?

It generates raster images and reference-guided variations by routing prompts to Codex, Pollinations or Gemini, writing a manifest with the run's parameters each time.

Can oma-image edit an existing image?

No — pixel-level editing, masking, cropping and format conversion are out of scope, and the skill redirects those requests to an image editor.

Source text

SKILL.md · MIT
---
name: oma-image
description: Generate raster images or reference-guided variations through the
  OMA image CLI. Use for image assets or explicit vendor comparisons; pixel
  editing uses an editor.
---

# Image Agent - Multi-Vendor Image Router

## Scheduling

### Goal
Generate images and visual assets through authenticated multi-vendor routing while preserving prompt clarity, reference-guided regeneration, cost controls, and reproducible output manifests.

### Intent signature
- User asks to generate images, visual assets, illustrations, product photos, concept art, mockups, or AI art.
- Another skill needs shared image-generation infrastructure.
- User provides reference images for a new generated variation or asks for vendor comparison.

### When to use

- Generating images, visual assets, illustrations, product photos, concept art
- Comparing output between multiple image models for the same prompt
- Producing images from prompts within editor workflows (Claude Code, Codex, Gemini CLI)
- Other skills needing image generation infrastructure (shared invocation)
- Regenerating a new image that follows a supplied reference's subject, style, lighting, or composition

### When NOT to use

- Pixel-level editing, masking, inpainting, object removal, compositing, cropping, resizing, or format conversion -> out of scope; use an image editor or a tool that explicitly supports that operation.
- Generating videos or audio -> out of scope
- Inline vector art / SVG composition from structured data -> use a templating skill
- Simple asset resizing or format conversion -> use a dedicated image library

### Expected inputs
- Image prompt or creative brief
- Optional vendor, size, quality, count, output directory, and reference images
- Authentication/environment state for Codex, Pollinations, or Gemini

### Expected outputs
- Generated image files under `.agents/results/images/` or requested output directory
- `manifest.json` with prompt, vendor, model, and reproducibility metadata
- Vendor comparison outputs when `--vendor all` is used

### Dependencies
- `oma image generate` CLI and vendor authentication
- Codex image generation, Pollinations API, or Gemini API/CLI strategy
- `resources/vendor-matrix.md`, `resources/prompt-tips.md`, and the `image:` section of `.agents/oma-config.yaml`

### Control-flow features
- Branches by prompt ambiguity, vendor auth, cost threshold, reference-guided regeneration support, path safety, and safety/timeout exit codes
- Calls external vendor APIs/CLIs
- Reads reference images and writes generated images plus manifests

## Structural Flow

### Entry
1. Validate that the request contains enough subject, setting, style, usage, and aspect-ratio signal.
2. Classify a supplied image as a reference for a newly generated variation, or as a request for unsupported pixel/mask editing.
3. Check authentication, cost guardrails, output path, and count limits.

### Transitions
- If the subject or intended result cannot be inferred, clarify that missing input; otherwise proceed with reasonable defaults.
- If `--vendor all` is requested, require every requested vendor to be available.
- If the request is reference-guided regeneration and the selected vendor supports references, pass it automatically.
- When the vendor is chosen by the agent (not fixed by an explicit `--vendor`), record it with subject `image.route-selection` per `references/_shared/runtime/route-selection.md`: vendor/model, and the auth, reference-support, or cost fact that decided it.
- If the request requires pixel, mask, crop, resize, or format editing, stop before generation and name the unsupported operation and an appropriate editor route.
- If estimated cost exceeds guardrail, require confirmation unless bypassed.

### Failure and recovery
- If auth is missing, report vendor-specific authentication requirement.
- If reference support is unavailable for the selected vendor, reject with actionable guidance.
- If local CLI is outdated, ask user to run `oma update`.
- If generation times out or is blocked, surface exit code and provider status.

### Exit
- Success: images and manifest exist in the output directory.
- Partial success: some vendors fail in comparison mode and failures are reported.
- Failure: no image is produced and the route/cost/auth/safety blocker is explicit.

## Logical Operations

### Tools and instruments
- `oma image generate`, `oma image doctor`, `oma image vendor list`
- Codex, Pollinations, and Gemini provider paths
- Prompt tips, vendor matrix, and image config

### Canonical command path
```bash
oma image doctor
oma image generate "<prompt>" --vendor auto --size auto --quality auto --output json
```

With reference images:
```bash
oma image generate --reference "<absolute-path>" --vendor codex "<prompt>"
```

### Resource scope
| Scope | Resource target |
|-------|-----------------|
| `LOCAL_FS` | Reference images, generated images, manifests |
| `PROCESS` | Provider CLIs and image router commands |
| `NETWORK` | Pollinations/Gemini or provider APIs |
| `CREDENTIALS` | Provider auth and API keys |

### Preconditions
- Prompt is sufficiently specified or user approves amplification.
- Required vendor auth and output permissions exist.
- Reference paths are accessible when used.

### Effects and side effects
- Creates image files and manifests.
- May call paid or rate-limited provider APIs.
- May read attached/reference images.

### Guardrails

1. **Resolve the brief**: use supplied details and reasonable creative defaults. Ask only when a missing choice materially changes the requested result. Do not require approval of an expanded prompt when generation is already authorized.
2. **Authentication-aware dispatch**: detect which vendor CLIs are available and run only those; with `--vendor all`, every requested vendor must be available (strict). Caveat: the `antigravity` health check verifies installation only (`agy --version`) — a signed-out agy passes health and fails at generate time with agy's own error.
3. **Cost guardrail**: confirm before executing runs whose estimated cost is ≥ `$0.20` (configurable). `--yes` / `OMA_IMAGE_YES=1` bypass. Choose generation from the requested outcome, not keyword presence. **Non-interactive contexts** (agents, CI — no TTY on stdin): the CLI cannot prompt, so a run at/over the threshold exits 1 with a message naming `--yes`. Use `--dry-run` to estimate cost. Reuse an existing budget authorization; otherwise confirm the additional spend before re-running with `-y`.
4. **Path safety**: output paths outside `$PWD` require `--allow-external-output`.
5. **Cancellable**: SIGINT/SIGTERM aborts in-flight provider calls and the orchestrator.
6. **Deterministic outputs**: every run writes `manifest.json` next to the images for reproducibility.
7. **Max `n` = 5**: wall-time bound.
8. **Exit codes align with `oma search fetch`** (0, 1, 2=safety, 3=not-found, 4=invalid-input, 5=auth-required, 6=timeout).

### Creative brief
Use the requested subject and constraints. Infer setting, style, lighting, and aspect ratio when unspecified; preserve explicit text and reference intent. Ask only for a material unresolved choice. Read `resources/prompt-tips.md` for an unfamiliar image category.

## References
- Provider invocation, references, and output layout: `resources/invocation.md` (selected vendor or reference-image operation only).

### Configuration

Project-specific settings: the `image:` section of `.agents/oma-config.yaml`, which `oma update` preserves. Shipped defaults live in the CLI (`DEFAULTS` in `cli/commands/image/config.ts`) — write only the keys you change. The legacy `config/image-config.yaml` is no longer read by the CLI; migration 022 moves anything you had changed there into oma-config (and deletes the file when it was never edited).
Env vars: `OMA_IMAGE_DEFAULT_VENDOR`, `OMA_IMAGE_DEFAULT_OUT`, `OMA_IMAGE_YES`, `POLLINATIONS_API_KEY`.

- Execution steps (follow for the selected task): `resources/execution-protocol.md`
- Vendor matrix: `resources/vendor-matrix.md`
- Prompt tips: `resources/prompt-tips.md`
- Checklist (run before handoff): `resources/checklist.md`
- Context loading: `references/_shared/core/context-loading.md`

© First Fluke · MIT · first-fluke/oh-my-agent

Similar skills

All skills for AI and ML →

oma-image by other authors comes first.

oma-image First Fluke oma-image dispatches AI image generation across Codex and Pollinations from your coding agent with auth checks and reproducible manifests. CX Codex +1 ★ 1,472repo stars seo-image-gen AgriciDaniel seo-image-gen — a skill for Claude Code, Codex, and Cursor builds search-optimized visuals via Gemini. Automates OG cards, blog hero banners, infographics, and clean product shots with preset aspect ratios. CX Codex +1 7,100installs videoagent-image-studio wells The videoagent-image-studio skill provides one-command access to Midjourney, Flux, Ideogram, and five other image models without storing API keys. The agent picks the best model and handles Midjourney's async workflow automatically. CX Codex +1 5,600installs blog-image AgriciDaniel Generates blog hero images, inline illustrations, and OG cards through paid Gemini API calls. Triggers on /blog image or when blog-write needs a cover. CX Codex +1 2,700installs generate-image K-Dense-AI The generate-image skill for Claude Code generates and edits images via OpenRouter Image API with Gemini, Recraft, and Seedream, for artwork, logos, and concept art. CX Codex +1 2,200installs ai-image-generator jezweb ai-image-generator — a skill that tells your agent which Gemini or Claude model to pick and how to build prompts in five parts. Useful when a site needs hero banners, icons or OG images. CX Codex +1 1,400installs scenario-gemini-image scenario-labs scenario-gemini-image enables image creation and reference-based editing with Google Gemini models via Scenario MCP. Handles multi-image fusion, video stills, and instruction edits. CX Codex +1 878installs

From the same set69

The whole set →
oma-pm oma-pm — a skill for Claude Code and Codex decomposes product requirements into scoped tasks with dependencies and acceptance criteria. Generates plan artifacts for specialist agents and orchestrators. CX Codex +1 ★ 1,472repo stars oma-design oma-design resolves a project's design direction and maintains DESIGN.md as the visual contract. Picks an external design skill — Interface Design, Taste, or Impeccable — for the task at hand. CX Codex +1 ★ 1,472repo stars oma-hwp oma-hwp converts Korean HWP, HWPX, and HWPML documents into Markdown or JSON for LLM prompts and retrieval pipelines. CX Codex +1 ★ 1,472repo stars oma-search The oma-search skill picks a search channel for docs, web research, code, or local project files, then labels external sources by domain trust. Use it when the right place to search is unclear. CX Codex +1 ★ 1,472repo stars oma-architecture oma-architecture is a Claude Code skill that helps design module and service boundaries, compare alternatives, and document architecture decisions as ADRs. CX Codex +1 ★ 1,472repo stars oma-coordination The oma-coordination skill guides manual coordination across specialist agents on multi-domain projects. It covers task planning, progress checks, contract alignment, and a final QA review. CX Codex +1 ★ 1,472repo stars oma-hwp oma-hwp turns Korean HWP, HWPX and HWPML documents into Markdown via kordoc. Extracts text, tables, lists, images and footnotes. CX Codex +1 ★ 1,472repo stars oma-pdf oma-pdf converts PDF documents into structured Markdown while preserving tables, headings, and lists. Runs when preparing files for LLM context or RAG pipelines. CX Codex +1 ★ 1,472repo stars

Project sponsors

Companies that keep this analytics open