agentleFS
Sign inStart free
ai-enhanced-engineerai-enhanced-engineer/aiee-skills/skills/ai-video-understanding/SKILL.md

ai-video-understanding

Skill4 starsChanged 5 months agoMITView on GitHub

Multimodal AI patterns for extracting structured information from video using Gemini, GPT-4o, and Claude vision. Frame sampling strategies, structured output with Pydantic, temporal reasoning, and cost optimization. Use for video analysis, report generation from video, and visual inspection.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this skill when building pipelines that extract structured information from video using multimodal LLMs, or when choosing between …

    line 12
  2. Distinctive:

    Use video-to-grid (48 frames in one image) for single-call analysis at ~$0.01/video

    line 35
  3. Distinctive:

    Use multi-pass: sparse low-res scan first, then dense high-res on flagged segments only

    line 38

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

AI Video Understanding

Use this skill when building pipelines that extract structured information from video using multimodal LLMs, or when choosing between native video input (Gemini) and frame-based approaches (GPT-4o, Claude).

Model Comparison

ModelVideo InputMax Duration/ImagesInput $/1M tokensBest For
Gemini 2.5 ProNative1 hour, 10 videos$4.00Long-form, temporal reasoning, cached Q&A
Gemini 2.5 FlashNativeSimilar to Pro~$0.50High-volume, cost-sensitive
GPT-4oFrame-basedHundreds of frames$2.50Structured extraction, schema adherence
Claude Sonnet 4.6Frame-based600 images/request$3.00High frame-count, multi-turn analysis

Frame Sampling Quick Reference

Task TypeRecommended FPSAccuracy
General overview0.5-1 FPS~85%
Action detection2-4 FPS~95%
Fine-grained motion8-16 FPS~98%+
Long-form (budget)Grid: 48 frames/image~70-80%

Cost Optimization

  • Enable Gemini context caching for multi-query workflows (90% discount on cached tokens)
  • Use video-to-grid (48 frames in one image) for single-call analysis at ~$0.01/video
  • Resize frames to 1024x1024 for optimal quality/cost balance
  • Apply transcript-first pattern: cheap audio analysis, then selective visual only where needed
  • Use multi-pass: sparse low-res scan first, then dense high-res on flagged segments only

See reference.md for detailed patterns and architecture decisions. See examples.md for production code examples.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free