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 agentmods add skills/devkindhq/ideogram-ai-toolkit/ideogram-promptnpx skills add devkindhq/ideogram-ai-toolkit --skill ideogram-promptgit clone --depth 1 https://github.com/devkindhq/ideogram-ai-toolkitWrote 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/devkindhq/ideogram-ai-toolkit/ideogram-prompt)<a href="https://agentmods.dev/skills/devkindhq/ideogram-ai-toolkit/ideogram-prompt"><img src="https://agentmods.dev/badge/skills/devkindhq/ideogram-ai-toolkit/ideogram-prompt.svg" alt="Measured on agentmods" height="20"></a>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 | $0.00128 | $0.00985 |
| Opus 5 | $0.00064 | $0.00492 |
| Sonnet 5 | $0.00026 | $0.00197 |
| Haiku 4.5 | $0.00013 | $0.00098 |
Grade A, and why
ideogram-prompt 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 4d 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 — 34 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Ideogram Prompting
Ideogram 4 was trained on structured JSON captions, not plain text — a plain-text prompt gets expanded into that structure by a "magic prompt" step before it ever reaches the model. That gives you two ways to prompt, and picking the right one depends on how much control the user actually wants.
Two modes
Loose / exploratory — write a natural-language prompt and let the model's own interpretation (magic prompt) fill in color, lighting, composition. Good for quick ideas, loose briefs, or when the user wants to be surprised. Just call mcp__ideogram__generate_image with a prompt string; don't over-specify.
Precise — when the user names an exact palette, a specific composition, or text that must render legibly, write the prompt as a structured caption instead of a vague adjective list. See references/json-caption-schema.md for the full schema (aesthetics/lighting/medium/color_palette, bounding-box elements, etc) — either follow its field structure in prose form, or paste the JSON itself into the prompt string as an experiment. Note: the connected generate_image tool has no explicit magic_prompt toggle (unlike edit_image, which does), so JSON-in-prompt isn't a guaranteed bypass the way it is in Ideogram's raw API — compare against a well-written prose version and keep whichever renders closer to what was asked for.
Either way, the highest-leverage lever is color_palette: up to 16 uppercase #RRGGBB hex codes steer the image's dominant colors directly, and up to 5 per element for per-subject control. If the user cares about exact colors, always name them as hex, not as color-adjectives ("teal" vs #0F766E). For controlled lighting, include both a highlight and a shadow hex, not just the dominant color — and if the background needs a specific tone, name that hex explicitly too rather than leaving it to inference.
Mechanically: sending a plain-text prompt to generate_image runs magic prompt (the model expands it into structure on your behalf); sending the JSON caption directly is closer to magic prompt off — what you wrote is closer to what renders. For plain-text prompts specifically, put the most important subject or action first — the model weights earlier tokens more heavily, so burying the actual subject after a paragraph of mood-setting adjectives works against you.
What ships with it
3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 4d ago First seen · 34 lines · 128 tokens per session scan A 837d549bb9c9
ideogram-prompt is a skill published in the GitHub repository devkindhq/ideogram-ai-toolkit (2 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 128 tokens to every session and 985 once invoked, about $0.0006 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
prompt-engineer
Writes, refactors, and evaluates prompts for LLMs — generating optimized prompt templates, structured output schemas, evaluation rubrics, and test suites. Use when designing prompts for new LLM applications, refactoring existing prompts for better accuracy or token efficiency, implementing chain-of-thought or few-shot…
prompt-master
Generates optimized prompts for AI tools. Activates only when the user explicitly asks to write, fix, improve, or adapt a prompt for a specific AI tool (LLM, Cursor, Midjourney, image AI, video AI, coding agents, etc.). Does not activate for general conversation, coding tasks, document writing, or other…
ai-workflow-architect
Designs AI systems, automations, and agent workflows for a business — identifying which manual work is worth automating, how to structure the system, which tools fit, and what could go wrong. Use this to automate part of an operation, design an agent or MCP workflow, reduce repetitive manual work, connect tools into a…
few-shot-examples
Curated few-shot examples for construction AI tasks: classification, extraction, analysis. Domain-specific examples for improved LLM performance.
cloudflare-workers-ai
Cloudflare Workers AI for serverless GPU inference. Use for LLMs, text/image generation, embeddings, or encountering AIERROR, rate limits, token exceeded errors.
prompt-templates
Reusable prompt templates for construction AI tasks: cost estimation, schedule analysis, document processing, BIM queries. Structured prompts for consistent results.