Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add hashiiiii/rules-for-ai --skill hashiiiii-imagesgit clone --depth 1 https://github.com/hashiiiii/rules-for-aiWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/hashiiiii/rules-for-ai/hashiiiii-images)<a href="https://agentmods.dev/skills/hashiiiii/rules-for-ai/hashiiiii-images"><img src="https://agentmods.dev/badge/skills/hashiiiii/rules-for-ai/hashiiiii-images/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/hashiiiii/rules-for-ai/hashiiiii-images"><img src="https://agentmods.dev/badge/skills/hashiiiii/rules-for-ai/hashiiiii-images.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00026 | $0.00887 |
| Opus 5 | $0.00013 | $0.00443 |
| Sonnet 5 | $0.00005 | $0.00177 |
| Haiku 4.5 | $0.00003 | $0.00089 |
Grade A, and why
hashiiiii-images scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 73 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Images
First, select vector or raster output. If words fully describe the image, write SVG code with the procedure below.
Do not make a flat geometric image as a raster image.
The process controls the image quality. Limit the design language. Then compare candidates at their target sizes before you select one.
When to Use
- Use this skill for logo marks, icons, favicons, promo tiles, and README header images.
- Use this skill for an OSS style with two to four shapes and one accent color.
If the image is illustrative or photorealistic, use a raster image generator. Then use a process that removes the background.
Gather Requirements First
Collect four inputs before you create an image. The request and the repository can already contain some inputs.
For each missing input, ask the user. Ask one question at a time. Include the default in each question.
Do not draw until you know all four inputs. A temporary accent or an exploratory draft still uses assumptions.
An incorrect background or color invalidates the complete contact sheet. One early question prevents work on incorrect candidates.
| Input | Confirm | Default to offer |
|---|---|---|
| Target | Identify the asset and its smallest usable size. | If the request does not specify a target, always ask. |
| Colors | Identify the accent color and the intended background. | Use a dark background and one accent color. |
| Motif | Identify the name meaning, domain concepts, and motifs to avoid. | Use the project name and README. |
| Output | Identify the file formats and target paths. | Create an SVG mark and PNG files at the target sizes. |
The target supplies the sizes for step 3. The motif supplies the metaphor list for step 2.
The Recipe
- Define the design language. Use two to four basic shapes for each mark. Use one accent color for each mark. Prefer negative space to additional detail. Do not use gradients or text.
- List metaphors. Write at least four visual metaphors from the project domain. Include the name meaning and the tool function. Give one metaphor to each candidate.
- Create a contact sheet. Put four to six candidate tiles in one SVG. Show each candidate at approximately 100 px. Also show it at its smallest target size. Use the intended background color. Render the SVG. Then inspect the PNG at its actual sizes.
- Select one candidate. Refine it with coordinate changes. If no candidate works, change the constraints. Then create a new contact sheet. Do not refine a weak candidate.
- Deliver only the foreground mark. Do not put a background tile in the final SVG. The context supplies the background. Use
fill-rule="evenodd"subpaths for transparent holes. Do not use shapes that have the background color.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago First seen · 73 lines · 26 tokens per session scan A 1d6fbe0dcf68
hashiiiii-images is a skill published in the GitHub repository hashiiiii/rules-for-ai (10 stars, last pushed 4d ago), licensed MIT. It adds 26 tokens to every session and 887 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
visual-asset-design
A guide for turning character, scene, and prop ideas into prompts for generating consistent reference images. It covers layouts, camera views, visual identity, and continuity across image variations.
diagram-generator
Generate a diagram and route to the right engine — draw.io XML (precise, editable, C4, swimlanes) or Excalidraw JSON (hand-drawn, sketch, wireframes). One entry for flowcharts, architecture, ER, sequence, mind maps. Don't use for Mermaid or slides.
excalidraw-visual-designer
Create and revise editable visual drawings directly in the Excalidraw website through the in-app browser. Use when the user asks Codex to draw in Excalidraw or update an Excalidraw canvas, including diagrams, flowcharts, architecture sketches, process maps, infographics, teaching visuals, product explanation graphics…
wechat-sticker-assets-designer
A design skill for creating the supporting images used when publishing WeChat sticker packs, including a banner, a support-request image, and a thank-you image.
lark-slides
A tool for creating and editing Feishu slide presentations, including their pages and content. Feishu is a workplace collaboration platform.
baoyu-infographic
A generator for infographics that combines one of 21 information layouts with one of 21 visual styles.