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.
git clone --depth 1 https://github.com/taotao135791-bit/gsap-transformWrote 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/rules/taotao135791-bit/gsap-transform/video-grammar)<a href="https://agentmods.dev/rules/taotao135791-bit/gsap-transform/video-grammar"><img src="https://agentmods.dev/badge/rules/taotao135791-bit/gsap-transform/video-grammar/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/rules/taotao135791-bit/gsap-transform/video-grammar"><img src="https://agentmods.dev/badge/rules/taotao135791-bit/gsap-transform/video-grammar.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 | $0.00095 | $0.00637 |
| Opus 5 | $0.00048 | $0.00318 |
| Sonnet 5 | $0.00019 | $0.00127 |
| Haiku 4.5 | $0.00010 | $0.00064 |
Grade A, and why
video-grammar 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 8d 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.
What it actually says
video-grammar
The aesthetic layer for video, beside motion-design-taste. motion-design-taste owns the page (dials, modes, type, color, layout); video-grammar owns what a page never has — shot framing, virtual camera movement, transitions, beat-synced pacing, aspect-ratio composition.
The .camera wrapper (the one idea)
All shot content lives inside one .camera wrapper; the master timeline animates the camera (scale/xPercent/yPercent), and the shot content animates inside it. Keep those two animation surfaces separate; mixing them creates the "everything moves at once" AI tell.
Shot types = camera scale
EWS 0.7–0.85 · WS 1.0 · MS 1.3 · MCU 1.6 · CU 2.2 · ECU 3.0–4.0. Offset with xPercent/yPercent to rule-of-thirds, not dead center every time.
Camera moves (one per shot — V1)
push (scale up) · pull (scale down) · pan (xPercent) · tilt (yPercent) · pedestal (y) · truck (x) · parallax-as-dolly (differential layer x) · orbit (rotationY — opaque bg only, V5). Ease power2.inOut/power3.inOut; none only for hard cuts; never back.*/elastic.* on a camera.
Transitions (one hero per video — V2)
hard cut (unlimited, default) · match cut (composition stays, content swaps) · crossfade · whip pan · morph · wipe · glitch. At most ONE signature transition per video; the rest are hard cuts.
BPM → beat-sync (V3)
beat = 60/bpm. Snap shot boundaries + hero reveals to beat boundaries. Bands: Restrained 60–80, Expressive 90–110, Cinematic 120–140 BPM.
Aspect safe areas (V4)
16:9 middle 80% h, avoid bottom 8%. 9:16 middle 60% v, top 12% + bottom 15% reserved. 1:1 centered 70%. Same scene must survive all three presets via .camera offset, not per-preset rewrites.
For full rules (V1–V5) and code patterns: see skills/video-grammar/SKILL.md.
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.
- 8d ago First seen · 30 lines · 95 tokens per session scan A 659c1c7d1993
video-grammar is a cursor rule published in the GitHub repository taotao135791-bit/gsap-transform (2 stars, last pushed 2mo ago), licensed MIT. It adds 95 tokens to every session and 637 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 2026-08-31.
Other cursor rules, from other repositories
080-creative-phase-metrics
description: WHEN entering a creative phase ENSURE comprehensive exploration and validation globs: ["/.md", "/.mdc"] alwaysApply: true.
livkiet-plugins-playai
Agent Framework plugin for voice synthesis with PlayAI API.
ponytail
Ponytail, lazy senior dev mode. Always pick the simplest solution that works.
angular-20
This rule provides comprehensive best practices and coding standards for Angular development, focusing on modern TypeScript, standalone components, signals, and performance optimizations.
dev-standard
Apache Superset development standards and guidelines for Cursor IDE.
cli-error-handling
CLI command error handling patterns.