Video Notes and Steps from Transcript and Frames
Timestamped notes and step lists from a video's transcript plus key frames, with visual gaps and caption errors flagged.
Before & after
Same prompt, same agent — the only difference is the skill.
Transcribing and summarizing a long video is definitely doable. To give you the best summary, I'd need to know a bit about the video—what's the topic, who's speaking, and what format would work best for you? I could grab the transcript if it's available on YouTube, but if there's no transcript, that's more complicated and might require external services. I can also pull some key frames if visuals are important, though that adds complexity. Let me know what you're working with, and I can suggest an approach. Do you have a preference for summary length or format—bullet points, narrative, or something else?
VIDEO PROCESSED: 90-min YouTube interview MODE USED: Efficient (40x speedup vs balanced mode) PROCESSING TIME: 87 seconds TOKEN COST: 2,100 tokens (vs 84,000 for full transcript) TRANSCRIPT STATUS: Built-in captions extracted FRAMES SELECTED: 12 (scene changes + deictic moments detected) KEY FINDINGS: SPEAKER: Dr. Jane Chen, AI researcher TOPIC: Scaling laws in neural networks—why bigger models get smarter TOP 3 INSIGHTS: 1. Scaling doesn't plateau (2:15-8:40 in video) → Prediction: 10x parameters = 15-20% performance boost → Implication: No evidence of diminishing returns yet 2. Training data quality now > quantity (23:45-31:20) → Key moment: "We wasted 2 years on raw internet scrapes" → Lesson: Curated datasets 3x more efficient 3. Emergent abilities appear without planning (52:15-59:30) → Models suddenly gain reasoning skills during scaling → Prediction paradox: Can't forecast which abilities emerge SCREEN SHARE MOMENTS: 3 detected (graphs at 34:50, live demo at 66:20, cited papers at 79:10) CITATIONS FOUND: → "Chinchilla scaling laws" (reference, year noted) → 3 arXiv papers mentioned (links auto-filed) FILE: ~/notes/chen-scaling-interview-summary.md FORMAT: Timestamped outline with embedded speaker quotes NEXT: Add to notion integration
About this skill
You don't have 40 minutes for that demo, and the auto-captions turn product names into nonsense and "click here" into a dead end. This skill turns a transcript plus a handful of frames into notes and step lists you can act on. It matches every on-screen moment to a frame and says plainly what was never seen.
What it does
- Says up front that a bare URL can't be watched, then gives the fastest way to get captions and frames, including commands that capture real frame timestamps
- Cleans rolling auto-captions so repeated phrases don't turn into fake quotes or pacing errors
- Pairs each "as you can see here" with the nearest matching frame, or marks it as a visual gap with its time range
- Settles caption-versus-screen conflicts: the frame wins for numbers, names, and code, and the speech wins for intent
- Writes steps as location, action, and success check, and lists prerequisites even when they're mentioned at the end of the video
- Flags misheard numbers, names, and CLI flags, plus outdated UI or version drift, for you to verify
What you get
Timestamped notes with the sources used, a TL;DR, prerequisites, ordered steps with frame evidence, visual gaps, and a verify list. A single question gets a single answer with a timestamp and a quote.
Best for
Engineers, ops staff, students, and analysts who turn software demos, tutorials, webinars, and lectures into notes or procedures without watching end to end.
SKILL.md preview
---
name: video-notes-and-steps-from-transcript-and-frames
description: Turns a video into trustworthy timestamped notes or ordered step lists. It reads the transcript or captions, aligns pointing phrases with key frames inside a time window, and flags caption errors and visual gaps instead of guessing. Use when the user says "summarize this video", "notes on this YouTube link", "I don't have time to watch this 40-minute demo", "pull the setup steps out of this screen recording", "can you watch this video?", "does the webinar mention pricing?", or "the auto-captions are garbage and I can't follow this tutorial".
---
# Video Notes and Steps from Transcript and Frames
… (sign up to view the full skill)Featured in
More productivity skills
View all Productivity skills →Apple-Style UI Engineer
Use when building any new UI component, screen, or page in React + Tailwind from a plain-text description, and the result must look like App
Zapier, Make and n8n Automation Recipe Builder
Designs no-code automations that run unattended: exact trigger, mapped fields, dedupe, error alerts, approvals, and monthly task cost.
AI Productivity System
A complete AI-driven system for planning your day, triaging tasks, and shipping more without burning out.
Daily Standup Writer from Git and Tickets
Turns your commits, PRs, reviews, and ticket moves into a 15-second standup post that never says shipped when it only merged.