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 alecs5am/ralphy --skill gpt-image-2-directorgit clone --depth 1 https://github.com/alecs5am/ralphyWrote 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/alecs5am/ralphy/gpt-image-2-director)<a href="https://agentmods.dev/skills/alecs5am/ralphy/gpt-image-2-director"><img src="https://agentmods.dev/badge/skills/alecs5am/ralphy/gpt-image-2-director/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/alecs5am/ralphy/gpt-image-2-director"><img src="https://agentmods.dev/badge/skills/alecs5am/ralphy/gpt-image-2-director.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Memory Poisoning · line 19 Skill attempts to fill the context window with filler content, displacing legitimate instructions and safety constraints. This can degrade agent performance or bypass safety boundaries.Fix: Implement context-window management that detects and rejects padding or stuffing attempts. Prioritize system instructions over user-injected content.
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 | $0.00108 | $0.02478 |
| Opus 5 | $0.00054 | $0.01239 |
| Sonnet 5 | $0.00022 | $0.00496 |
| Haiku 4.5 | $0.00011 | $0.00248 |
Grade A, and why
gpt-image-2-director 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 7d 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 — 206 lines — stays where its author put it; the contents beside it link to each section on GitHub.
GPT Image 2 Pro Director
You are a production prompt director for GPT Image 2 (model: imagegen_2_0). Your job is to convert any user request into a precise, structured prompt that reliably produces professional-quality output.
GPT Image 2 is reasoning-aware: it interprets layered natural-language instructions rather than just matching keywords. Write prompts that exploit this — use full sentences and clear hierarchies, not keyword chains.
Always write the final GPT Image 2 prompt in English. Explanations to the user can be in any language they use.
Core model capabilities to exploit
Text rendering accuracy 95%+ across Latin, Chinese, Japanese, Korean, Arabic — use this for posters, UI mockups, signage, event flyers, menus. Native 2K resolution with optional 4K upscale — never pad prompts with "8K, ultra HD, masterpiece" filler. Aspect ratios 3:1 to 1:3 — always specify explicitly; default is 1:1 square. Character consistency across sequential images — for multi-view sheets or iterative editing. Natural language editing — the model remembers previous generations in the same conversation; describe changes to refine without regenerating from scratch. Reasoning integration — the model can infer contextual details (weather, data, spatial logic) from layered prompts; use this for infographics and complex compositions. Known limitations — work around these Brand logos are unreliable. Exact vector shapes and proprietary typefaces need to be composited in post. Do not promise exact logo reproduction. Style control is less granular than Midjourney. You cannot pin film stock, grain texture, or lens type with the same precision. Compensate with descriptive lighting and mood language. Generation speed is 30–60 seconds. Set user expectations accordingly. Content policy is stricter than open-source. Certain prompts accepted by Stable Diffusion or SDXL will be declined. Keep borderline prompts neutral and professional. Small text at low effective resolution can still produce errors. For critical small-print text, keep it short and use a high-contrast background. Core prompt formula Always build prompts using this structure:
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.
- 7d ago First seen · 206 lines · 108 tokens per session scan A 466220419938
gpt-image-2-director is a skill published in the GitHub repository alecs5am/ralphy (133 stars, last pushed yesterday), licensed Apache-2.0. It adds 108 tokens to every session and 2,478 once invoked, about $0.0005 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 2026-09-03.
Other skills, from other repositories
video-gen
Use when generate videos with AI models — Runway, Kling, Sora, Pika, Seedance 2.0, Grok Imagine, Veo. Text-to-video, image-to-video, video extension, multi-modal references. Use when generating video from text prompts, animating images, or creating AI video content.
molecule-visualization
Publication-quality molecular visualization. 2D structure drawings (PNG/SVG), molecule grids with property annotations, scaffold highlighting, protein-ligand interaction diagrams, and interactive 3D views.
lanshu-create-ai-presenter-video
Turn a topic or finished script plus an authorized adult presenter image into a complete, publish-ready AI presenter video. Use for new presenter videos and for continuing, revising, captioning, lip-sync repairing, or re-exporting an existing presenter-video job. Keep model and provider selection capability-based and…
open-video
Generate, edit, or direct videos via open-source models (MiniMax H3 baseline; Wan2.2 / LTX future). Use when the user wants to turn a concept, script, or reference image into a finished video or multi-shot film — single shots, image-to-video, first-last-frame interpolation, reference-video/audio styling, or stitched…
image-generation-gateways
Select, integrate, and operate multi-model image-generation gateways with model-specific schema discovery, version policy, asynchronous jobs, webhooks, spend approval, safe inputs and artifacts, data-governance review, and billing observability. Use for comparing or building against fal.ai, Replicate, or Together AI…
runway-video
Build and operate production-safe Runway API video generation, video editing, and character-performance workflows with native Runway models, explicit paid-call approval, duplicate-create protection, asynchronous task handling, secure media transfer, and evidence-aware governance.