image-generation
Production image generation through Modal + ComfyUI or configured providers.
Skill by bl1nk42
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/bl1nk42-image-generation
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: writing
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md and code. Checked Oct 26, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/image-generation
What's inside
- Skill name
image-generation- Description
- Production image generation through Modal + ComfyUI or configured providers. Use when creating, editing, analyzing, caching, queueing, or registering images, including text-bearing visuals, references, style consistency, and multi-image batches.
- License
- Apache-2.0
- Instructions
- 34 lines in SKILL.md · view on GitHub
Files (5)
- SKILL.md
- batch-and-queue.md · references/
- model-routing.md · references/
- prompt-contract.md · references/
- quality-gates.md · references/
SKILL.md (first lines)
# Image Generation ## Mission Deliver a usable image, not merely a prompt. Choose the lowest-cost route that satisfies the request, preserve user constraints, make the executed prompt inspectable, and avoid duplicate generation. ## Route first 1. Classify the job: new image, edit, restoration, precise layout, meme, character continuity, or video reference. 2. Choose the provider/model from `model-catalog.json`; check `supports`, aspect ratios, edit capability, and queue availability. 3. If the request is a diagram or numeric chart, do not use generative imagery as the source of truth: use Mermaid or deterministic plotting. 4. If an image is supplied, call `analyze_image` before editing unless the requested change is unambiguous and reversible. 5. Normalize prompt with `prepare_prompt_tool`, then check cache using the full request signature. ## Prompt contract Structure the prompt as: purpose/medium, subject, composition, lighting/material, exact text, constraints, avoid list. Preserve names, numbers, dates, and requested language exactly. Separate subject facts from learned style. Apply style suffix only after the subject is complete. Negative prompts must name bare concepts, not sentences such as “do not”. ## Execution - One image: call the provider after cache lookup. - Multiple independent images: use `queue_image`; retain every `job_id`; use `get_job` rather than resubmitting. - Dependent sequence: execute sequentially and pass the actual prior output as reference. - Edit: verify `supports.edit`; describe only the requested change and explicit preservation constraints. - After success: return image, model, executed prompt, aspect ratio, cache state, job ID if any, and registration status. ## Quality gate Pass only if the output has the requested subject, composition/aspect, text or no-text constraint, reference fidelity, and no fatal artifacts. Regenerate once with one targeted correction. Do not silently make endless aesthetic refinements. ## Privacy and memory Do not auto-save private images as references. Save to registry only when requested or when the user explicitly opts into learning. Use `save_as_reference`, `register_entity`, and `vote_entry` to make preference changes explicit. ## Read when needed - `references/prompt-contract.md` for templates and structured output. - `references/model-routing.md` for provider/model decisions. - `references/quality-gates.md` for scenario checks. - `references/batch-and-queue.md` for concurrency and retries.
From bl1nk42/image-gen, Apache-2.0.
More from bl1nk42/image-gen
- 0.600score
photo-restoration
bl1nk42
Restore blurry, damaged, faded, or vintage photos while preserving identity, pose, composition, clothing, and historical character.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
nothing-design
bl1nk42
Apply the Nothing-inspired monochrome industrial design system to interfaces, dashboards, image prompts, and generated visual references.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
video-generator
bl1nk42
Plan and execute professional AI video projects including commercials, explainers, short films, social clips, and image-to-video sequences.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
memory-recall
bl1nk42
Retrieve relevant project decisions, prior prompts, style preferences, debugging notes, and past outputs with source traceability.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
prompt-engineer
bl1nk42
Design, refactor, evaluate, and version prompts for image generation, image analysis, style learning, and structured agent tools.
SkillAFreshness: A★ 2Claude Code +9 - 0.595score
character-story-video
bl1nk42
Plan and produce a multi-part animated character story with stable identity, sequential scenes, reference images, image-to-video clips, audio planning, and…
SkillAFreshness: A★ 2Claude Code +9 - 0.595score
meme-generator
bl1nk42
Create captioned memes from classic templates or a supplied background, with correct text encoding, line limits, preview, and export.
SkillAFreshness: A★ 2Claude Code +9 - 0.593score
visual-production-router
bl1nk42
Route visual requests to image generation, restoration, Nothing design, memes, character stories, video production, prompt engineering, memory, or the exact…
SkillAFreshness: A★ 2Claude Code +9 - 0.590score
style-learning
bl1nk42
Learn and improve reusable image styles from liked references, votes, prompts, embeddings, and recent generation outcomes.
SkillAFreshness: A★ 2Claude Code +9 - 0.590score
dataset-curation
bl1nk42
Curate generated images into a clean training dataset, validate metadata, remove duplicates, and sync manifests/assets to Hugging Face.
SkillAFreshness: A★ 2Claude Code +9 - 0.587score
modal-operations
bl1nk42
Deploy, inspect, warm, cache, and troubleshoot the Modal + ComfyUI image backend and Hugging Face storage.
SkillAFreshness: A★ 2Claude Code +9
Signals
- Stars
- 2
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Sep 25, 2026 (1 month ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- Apache-2.0
- Repository
- github.com/bl1nk42/image-gen
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from bl1nk42
- 0.600score
photo-restoration
bl1nk42
Restore blurry, damaged, faded, or vintage photos while preserving identity, pose, composition, clothing, and historical character.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
memory-recall
bl1nk42
Retrieve relevant project decisions, prior prompts, style preferences, debugging notes, and past outputs with source traceability.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
nothing-design
bl1nk42
Apply the Nothing-inspired monochrome industrial design system to interfaces, dashboards, image prompts, and generated visual references.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
prompt-engineer
bl1nk42
Design, refactor, evaluate, and version prompts for image generation, image analysis, style learning, and structured agent tools.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
video-generator
bl1nk42
Plan and execute professional AI video projects including commercials, explainers, short films, social clips, and image-to-video sequences.
SkillAFreshness: A★ 2Claude Code +9 - 0.595score
character-story-video
bl1nk42
Plan and produce a multi-part animated character story with stable identity, sequential scenes, reference images, image-to-video clips, audio planning, and…
SkillAFreshness: A★ 2Claude Code +9
Cite this page
image-generation skill. Agent Mart. https://agentmart.dev/skills/bl1nk42-image-generation (last updated 2026-10-24)