ahmedsakri/youtube-agent-pro / skills/yt-chapters
yt-chapters
VerifiedDraft YouTube chapters from a timestamped transcript and check basic timestamp constraints before the creator uses them. Use for "add chapters", "timestamps", "break this video into sections".
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 1.4 KB |
| chapters.py | script | 3.8 KB |
chapters.py
View on GitHub#!/usr/bin/env python3
"""chapters.py - YouTube chapters from a timestamped transcript.
python3 chapters.py transcript.srt # or .vtt / whisper .json
python3 chapters.py transcript.srt --target 8 --json
Prints a description block you can paste straight under a video. YouTube's own rules, enforced here
rather than assumed: the list must start at 00:00, needs at least three entries, and each chapter
must be at least 10 seconds long. A block that breaks any of those silently does not become
chapters, which is why this checks instead of trusting.
Boundaries come from the gaps - the pauses you actually took between sections - scored by how long
the pause was and how much the vocabulary changes across it. It is a first draft you retitle, not a
summariser.
"""
import argparse, json, os, re, sys
sys.path.insert(0, os.path.join(os.path.dirname(os.path.abspath(__file__)), "..", "yt-edit"))
from deadair import load, parse_ts # noqa: E402 (same parser, one implementation)
STOP = set("the a an of for to in on and or is are was were be been with this that it as at by from "
"you your i my we our they them he she but so if then than there here what which who how "
"when where why not no yes do does did just really very like about into over out up down "
"can could will would should have has had get got make made go going went one two".split())
def keywords(text):
return {w for w in re.findall(r"[a-z']{4,}", text.lower()) if w not in STOP}
def mmss(t):
t = int(t); h, m, s = t // 3600, (t % 3600) // 60, t % 60
return f"{h}:{m:02d}:{s:02d}" if h else f"{m}:{s:02d}"
def main():
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument("transcript")
parser.add_argument("--target", type=int, default=7)
parser.add_argument("--json", action="store_true")
args = parser.parse_args()
target, as_json = args.target, args.json
if target < 3:
parser.error("--target must be at least 3")
try:
cues = load(args.transcript)
except (OSError, ValueError, KeyError, TypeError) as exc:
parser.error(str(exc))
if len(cues) < 6: parser.error("too few cues to chapter")
dur = max(c[1] for c in cues)
cand = []
for i in range(1, len(cues)):
gap = cues[i][0] - cues[i - 1][1]
before = " ".join(c[2] for c in cues[max(0, i - 12):i])
after = " ".join(c[2] for c in cues[i:i + 12])
kb, ka = keywords(before), keywords(after)
shift = 1 - (len(kb & ka) / len(kb | ka)) if (kb | ka) else 0
cand.append((gap * 1.6 + shift * 3.2, cues[i][0], i))
cand.sort(reverse=True)
picked, MIN = [0.0], 10.0
for _, t, i in cand:
if len(picked) >= target: break
if all(abs(t - p) >= MIN for p in picked) and dur - t >= MIN:
picked.append(t)
picked.sort()
chapters = []
for n, t in enumerate(picked):
end = picked[n + 1] if n + 1 < len(picked) else dur
text = " ".join(c[2] for c in cues if c[0] >= t and c[1] <= end)
kw = [w for w in keywords(text)]
kw.sort(key=lambda w: (-text.lower().count(w), w))
title = " ".join(w.capitalize() for w in kw[:3]) or "Section"
chapters.append({"start": round(t, 2), "label": mmss(t), "draft_title": title,
"seconds": round(end - t, 2)})
ok = len(chapters) >= 3 and chapters[0]["start"] == 0 and all(c["seconds"] >= MIN for c in chapters)
if as_json:
print(json.dumps({"valid": ok, "chapters": chapters}, indent=1)); return
print()
for c in chapters: print(f" {c['label']} {c['draft_title']}")
print(f"\n {len(chapters)} chapters"
f"{'' if ok else ' -- INVALID: YouTube needs 3+, a 00:00 first entry and 10s minimum each'}")
print(" Retitle every line before pasting. These are the topic words, not your words.\n")
if __name__ == "__main__":
main()
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |