Cassiiopeia/projectops / skills/pro-image
pro-image
VerifiedMake image assets (backgrounds, characters, icons, illustrations, items) that fit the project's own style: you write the prompt from a per-repo harness, Codex or Antigravity draws it, the CLI fixes size and transparency. Use for 'make an image', 'draw', 'generate an icon'.
references/style-explore.md
View on GitHub# Style explore: let the user see and pick the look
> When to read: the repo has no `STYLE.md`, the user asks "what style fits?", or a new project starts its art.
> People only know which look fits their project when they see it next to others.
## Flow
1. **Read the project's mood.** App name, audience, screens, colors in the design tokens, any existing art and
copy tone. Write one line: "a calm seaside visual novel for adults", "a playful SQL quiz for students".
2. **Pick 3-4 distinct looks** that could fit that mood, and one subject that shows them well (the main character,
the app mascot, or the most-used background). Make them really different, e.g.:
- anime watercolor with thin ink lines
- flat vector with bold shapes and 3 colors
- soft 3D clay
- pixel art 32 px
- hand-drawn pencil and marker
- paper cut-out collage
3. **Generate each look** with the same subject, same size, same provider, in the background, one by one
(`run --name style_<n> ...`). Only the style lines differ between the prompts.
4. **Show one board.** `board style_1.final.png style_2.final.png ... --out style_board.png`, open it with Read to
check it, show it to the user, and ask with your agent's choice prompt (AskUserQuestion in Claude Code; a numbered list elsewhere): one option per number with a 3-6 word name of the
look, plus "mix: describe it". Do not show the four runs separately and do not describe the process.
5. **Lock the pick.** Write `STYLE.md` from the chosen prompt's style lines (concrete: line, shading, palette with
hex, light), copy the chosen image to `refs/`, `record --status approved`, and `record --status rejected` for the
others with the user's reason if they gave one.
6. **Default when the user skips:** pick the look that matches the existing app best (or the first one), say which
in one line, and continue. They can run this again any time.
Run it again when the user wants to change direction ("make it cuter", "try other styles"): same subject, new looks,
and include the current look as option 1 so they can keep it.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |