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 chadixearth/graphyloop --skill ai-video-prompt-engineergit clone --depth 1 https://github.com/chadixearth/graphyloopWrote 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/chadixearth/graphyloop/ai-video-prompt-engineer)<a href="https://agentmods.dev/skills/chadixearth/graphyloop/ai-video-prompt-engineer"><img src="https://agentmods.dev/badge/skills/chadixearth/graphyloop/ai-video-prompt-engineer/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/chadixearth/graphyloop/ai-video-prompt-engineer"><img src="https://agentmods.dev/badge/skills/chadixearth/graphyloop/ai-video-prompt-engineer.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00077 | $0.01771 |
| Opus 5.5 | $0.00031 | $0.00708 |
| Sonnet 5.5 | $0.00015 | $0.00354 |
| Haiku 4.5 | $0.00008 | $0.00177 |
Grade A, and why
ai-video-prompt-engineer 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 21d 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 — 140 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AI Video Prompt Engineer
Overview
Generate production-quality AI video prompts through an interactive 7-question flow. Output is always a single unified prompt describing all segments in one continuous timeline — one copy-paste, no repeated pasting per segment. Designed for OmniFlash and similar AI video models.
The Output Contract
The output file MUST follow this exact structure:
# [Title] — [Duration] [Style] Film
## Single Unified Prompt — One Paste for OmniFlash
---INSTRUCTION (read first)---
Rules the AI video model must follow:
- Anti-Slop Constitution: no glitch, no lens flare, no floating AI brains, no robots, no particle explosions, no hexagonal wireframe tunnels, no holographic hands
- Real humans only — every scene features people doing real work (no androids, no synthetic humans, no faceless silhouettes)
- Natural color names only in prompt body — hex codes go in FOR EDITOR section at bottom
- Mobile-readable text overlays (minimum 5% of frame height), clean geometric sans-serif, high contrast
- Natural motivated lighting (window light, monitor glow, golden hour) — not flat studio
- Smooth purposeful camera movement (gimbal/dolly/slider) — no handheld/shaky
- Voice-visual match: tone of visuals matches tone of the brand
---[00:00-00:XX] SEGMENT 1 — NAME---
Full scene description for segment 1. Include:
- Camera movement, lighting, color palette, human activity, setting
- Text overlay content, timing, position, color (natural name)
- Transition instruction for end of segment
---[00:XX-00:YY] SEGMENT 2 — NAME---
Full scene description for segment 2. Same level of detail. No cross-references to segment 1 — fully standalone description.
---[00:YY-00:ZZ] SEGMENT 3 — NAME---
Full scene description for segment 3.
---FOR EDITOR — Hex Codes & Technical Specs---
| Natural Name | Hex Code | Usage |
|---|---|---|
| [description] | #[code] | [where used] |
Plus: font specs, logo rules, aspect ratio, audio direction.
Interactive Flow
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.
- 21d ago Changed d15d38fde1d6
- 2mo ago First seen · 140 lines · 77 tokens per session scan A 2656703d16f2
ai-video-prompt-engineer is a skill published in the GitHub repository chadixearth/graphyloop (3 stars, last pushed yesterday), licensed MIT. It adds 77 tokens to every session and 1,771 once invoked, about $0.0003 per session on Opus 5.5. 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
showrunner
Conduct a product from the owner's idea to release through one enforced lifecycle (setup, discovery, assessment, constitution, business documents, then per initiative intake, roadmap, spec, design, design review, handoff, Arc plan, Step 0, build, verify, security, acceptance, merge, release, close) using the Forge…
sentry
Sweep, triage, test, fix, verify, and maintain software security posture with categorical coverage, governed penetration testing, durable accepted-risk memory, refreshable standards knowledge, isolated fixes, and an unconditional stop before main. Use when initializing security bindings, reviewing application, mobile…
forge
Conceive products and features with an inventor without silently making product decisions. Use when initializing product bindings, discovering a product constitution from a raw idea, planning phases, writing a decision-complete feature specification, or appending an evidence-backed product decision before…
pen-test
Plan, execute, validate, retest, and report authorized penetration tests under Sentry with explicit rules of engagement, safety-classed actions, versioned test-case coverage, reproducible evidence, independent finding validation, and cleanup. Use when testing an owned or explicitly authorized web, API, mobile, cloud…
arc
Plan, dispatch, verify, and merge isolated software implementation arcs with maximum safe autonomy and an unconditional stop before main. Use when initializing repository build bindings, consuming an approved Forge specification or brief, rendering an implementer prompt, running a cold-context worker in an isolated…
bible
Synthesize the read-only architecture and capabilities document ("Bible") from the repository, the Forge spec, Arc ship reports, and Sentry security artifacts, binding every claim to a real path, interface, or named report and routing spec-promised but unbuilt capabilities to intent-not-shipped. Use when binding Bible…