ahmedsakri/youtube-agent-pro / openai/skills/yt-idea
yt-idea
VerifiedTurn a niche or a blank page into video ideas that are searchable, specific and bigger than the channel - each one scored before it costs a week of filming. Use for "what should I make", "I have no idea what to post", "give me video ideas", "ideas for my channel", or when a creator is staring at an empty upload schedule.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 2.4 KB |
| agents/openai.yaml | config | 244 B |
| ideascore.py | script | 5.9 KB |
ideascore.py
View on GitHub#!/usr/bin/env python3
"""ideascore.py - score a video idea before you spend a week shooting it.
python3 ideascore.py ideas.txt # one idea per line, ranked
python3 ideascore.py --idea "one line" # score a single idea
python3 ideascore.py ideas.txt --json # machine-readable
An idea is scored on four axes, each 0-100, and a verdict that is 60% the mean and 40% the weakest
axis. The weakest-link weighting is the same choice hookscore.py makes: an idea that is searchable,
specific and broad but has no stake is an idea nobody clicks, and averaging hides the dead axis.
SEARCH does it answer a query a real person types? question words, a concrete noun, intent.
SPECIFIC a number, a name, a timeframe - or an adjective doing the work a fact should.
REACH is the audience bigger than you? "for react devs" is narrow, "for anyone" is nobody.
DURABLE evergreen earns views for years; a dated take earns them for a week. Neither is wrong,
but you should know which one you just greenlit.
WHAT THIS IS NOT. It does not know your niche, your competition, or whether the idea has been done.
It reads the words of the idea, not the market. A high score means the idea is well-shaped; it is
not a demand forecast. Pair it with yt-viral, which reads what is actually working.
"""
import argparse, json, os, re, sys
QUESTION = re.compile(r"\b(how|why|what|which|when|where|should|can|is|are|do|does|will)\b", re.I)
INTENT = re.compile(r"\b(guide|tutorial|review|vs|versus|best|fix|stop|avoid|build|make|setup|"
r"explained|mistakes|beginners?|without|using|in \d+)\b", re.I)
VAGUE = {"amazing","incredible","insane","crazy","huge","massive","ultimate","best","powerful",
"secret","revolutionary","epic","perfect","everything","stuff","things","content"}
NUMBER = re.compile(r"\b(\d[\d,.]*\s?(%|k|m|x|hrs?|hours?|mins?|minutes?|days?|weeks?|months?|years?|dollars?|\$)?)\b", re.I)
TIMELY = re.compile(r"\b(202\d|today|this week|this month|right now|new|latest|just|update|"
r"breaking|announced|released|beta|v\d)\b", re.I)
EVERGREEN = re.compile(r"\b(how to|guide|explained|beginners?|basics|fundamentals|101|complete|"
r"ever|always|principles?|mistakes)\b", re.I)
NARROW = re.compile(r"\bfor (react|vue|rust|haskell|niche|enterprise|senior|advanced)\b", re.I)
TOO_BROAD = re.compile(r"\bfor (everyone|anyone|all|people)\b", re.I)
def words(t):
return re.findall(r"[a-z0-9'$%]+", t.lower())
def search(t):
s = 20
if QUESTION.search(t):
s += 26
s += min(30, 15 * len(INTENT.findall(t)))
# a bare topic with no verb is a title, not a query
if not re.search(r"\b(how|why|what|guide|review|vs|fix|build|make|stop)\b", t, re.I):
s -= 12
return max(0, min(100, s))
def specific(t):
w = words(t)
if not w:
return 0
nums = len(NUMBER.findall(t))
vague = sum(1 for x in w if x in VAGUE)
proper = sum(1 for x in t.split()[1:] if x[:1].isupper() and len(x) > 2)
return max(0, min(100, 32 + nums * 20 + proper * 8 - vague * 18))
def reach(t):
s = 55
if TOO_BROAD.search(t):
s -= 30 # "for everyone" addresses nobody
if NARROW.search(t):
s -= 10 # narrow is fine, just capped
if re.search(r"\bfor \w+", t, re.I):
s += 15 # names an audience at all
return max(0, min(100, s))
def durable(t):
ever = len(EVERGREEN.findall(t))
timely = len(TIMELY.findall(t))
return max(0, min(100, 50 + ever * 16 - timely * 20))
AXES = [("SEARCH", search), ("SPECIFIC", specific), ("REACH", reach), ("DURABLE", durable)]
FIX = {
"SEARCH": "frame it as the question someone types - start with how/why/what",
"SPECIFIC": "put a number, a name or a timeframe in it",
"REACH": "name one audience, not 'everyone' and not a sub-niche of a sub-niche",
"DURABLE": "if it is a timely take, say so - it will earn for a week, not a year",
}
def score(t):
parts = {n: fn(t) for n, fn in AXES}
vals = list(parts.values())
verdict = round(0.6 * (sum(vals) / len(vals)) + 0.4 * min(vals))
life = "timely" if durable(t) < 45 else "evergreen" if durable(t) > 62 else "mixed"
return parts, verdict, life
def band(v):
return "STRONG" if v >= 70 else "WORKABLE" if v >= 52 else "WEAK"
def report(t, parts, verdict, life):
print(f"\n {t.strip()}")
print(f" {'-' * min(72, max(20, len(t.strip())))}")
for k, v in parts.items():
print(f" {k:<10} {v:3d} {'#' * (v // 5)}")
print(f" {'VERDICT':<10} {verdict:3d} {band(verdict)} ({life})")
low = min(parts, key=parts.get)
print(f" weakest {low} - {FIX[low]}")
def main():
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument("file", nargs="?")
parser.add_argument("--idea")
parser.add_argument("--json", action="store_true")
args = parser.parse_args()
as_json = args.json
if (args.file is None) == (args.idea is None):
parser.error("provide a text file or --idea, not both")
try:
lines = [args.idea] if args.idea is not None else open(args.file, encoding="utf-8").read().splitlines()
except OSError as exc:
parser.error(str(exc))
lines = [line for line in lines if line.strip()]
if not lines:
parser.error("provide at least one nonempty idea")
out = []
for t in lines:
parts, verdict, life = score(t)
out.append({"idea": t.strip(), "axes": parts, "verdict": verdict,
"band": band(verdict), "life": life})
out.sort(key=lambda r: -r["verdict"])
if as_json:
print(json.dumps(out, indent=1))
return
for r in out:
report(r["idea"], r["axes"], r["verdict"], r["life"])
if len(out) > 1:
print(f"\n pick: {out[0]['idea'].strip()} ({out[0]['verdict']}, {out[0]['band']})\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: 194k | 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: 194k | 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: 194k | Updated | Claude CodeCodex | Verified | Install |