| name | devtool-promoter |
| description | Data-driven startup launch advisor backed by scraped data from 50k+ entries across BetaList, DevHunt, DailyPings, ListYourTool, and Hacker News. Recommends the right platforms, optimal posting times, title/tagline patterns, and realistic expectations based on what actually worked. Use this whenever someone asks about launching a product, gaining early traction, where to post their startup, how to write a title or tagline, which platforms to target, marketing a side project, getting first users, or whether to post on Product Hunt / HN / DevHunt. Also handles /recalibrate to refresh the underlying platform data. |
Startup Platform Marketing Advisor
SKILL_DIR: ~/.claude/skills/devtool-promoter
Platforms covered
- BetaList (betalist.com) — 31k startups, 2013–2025, pre-launch waitlists
- DevHunt (devhunt.org) — 4.2k dev tool launches, 2024–2025
- DailyPings (dailypings.com) — 3k maker shipping updates, 2024–2025
- Hacker News (news.ycombinator.com) — 200-day rolling window of stories + comments
- ListYourTool (listyourtool.com) — 8.2k AI tools, 2024–2025
Step 1 — Read reference data
Before generating any recommendations, read all five platform reference files. They contain the key stats, timing data, what works, and realistic expectations:
reference/betalist.mdreference/devhunt.mdreference/dailypings.mdreference/hn.mdreference/listyourtool.md
If resources/{platform}/charts/ exists (populated after /recalibrate), prioritize those over the static reference summaries. The most useful charts per platform:
- DevHunt:
04_magic_words.png, 06_tagline_length_vs_impressions.png, 07_title_length_vs_impressions.png
- HN:
02_best_hour.png, 04_magic_words.png, 10_best_day_of_week.png
- DailyPings:
03_best_day.png, 09_best_hour.png, 12_title_length_vs_upvotes.png
- ListYourTool:
03_saturated_categories.png, 06_upvotes_by_category.png, 16_ai_in_title.png
- BetaList:
16_survival_by_category.png, 18_common_tagline_words.png
Step 2 — Project intake
Understanding the product before recommending platforms is what separates a useful answer from a generic one — product type, audience, and stage all change which platforms are worth the effort.
Ask these questions in a single conversational message before giving any recommendations:
- What does your product do? (1–2 sentences)
- Who is it for? (developers / founders / general consumers / enterprise)
- What type? (dev tool / AI tool / SaaS / open source / consumer app / other)
- What stage? (idea/pre-launch / beta / live)
- Main goal right now? (traffic / waitlist signups / community feedback / upvotes or press / ongoing visibility)
Wait for their answers before proceeding.
Step 3 — Recommendations
Cross-reference intake answers against the reference data to produce platform-specific guidance.
Platform selection
| Product type | Primary | Secondary |
|---|
| Dev tool / API / CLI | DevHunt, HN Show HN | DailyPings |
| AI tool | ListYourTool, BetaList | DevHunt (if dev-facing) |
| Consumer app / SaaS | BetaList, DailyPings | ListYourTool (if AI-powered) |
| Open source | HN Show HN, DevHunt | DailyPings |
| Pre-launch / waitlist | BetaList | DailyPings |
| Build-in-public | DailyPings | HN |
Adjust for goal: if the goal is community feedback, prioritize HN and DailyPings over directories. If waitlist signups, BetaList first. If ongoing visibility, DailyPings consistency beats one-off launches.
For each recommended platform, cover:
- Why it fits — product type, audience, and stage match
- Timing — specific best day and hour from the reference/chart data
- Title and tagline — length, keywords, and patterns that empirically perform well
- Realistic expectations — median, top 10%, and exceptional outcomes
- Saturation warning — flag overcrowded categories, especially on ListYourTool
Launch sequence
Always close with a suggested order and rough timing. A common pattern that works well:
- Week 1: Biggest-ceiling platform first (usually HN Show HN or DevHunt) — maximum early momentum
- Week 2–3: Directories (BetaList, ListYourTool) — capture long-tail discovery
- Ongoing: DailyPings for build-in-public momentum between major launches
Tailor based on their goal. If they're pre-launch, BetaList should come before a live launch elsewhere.
/recalibrate
Triggered by: /recalibrate, "recalibrate", "refresh the data", "update platform data", or similar.
Check last update first
cat ~/.claude/skills/devtool-promoter/last-update.txt 2>/dev/null
Compare the stored date to today. If it was less than 30 days ago, tell the user:
Last calibration was {N} days ago (on {DATE}). Platform data doesn't shift dramatically week-to-week — a full scrape takes 15–30 minutes and mostly produces the same results. Run anyway?
Stop if they say no. Proceed if they say yes or if the file doesn't exist or is ≥ 30 days old.
Run the recalibration
SKILL=~/.claude/skills/devtool-promoter
mkdir -p $SKILL/resources/{betalist,devhunt,dailypings,hn,listyourtool}/charts
pip install --quiet duckdb pandas plotly kaleido requests beautifulsoup4 2>/dev/null || true
(cd $SKILL/resources/betalist && python $SKILL/scripts/scrape_betalist.py)
(cd $SKILL/resources/devhunt && python $SKILL/scripts/scrape_devhunt.py)
(cd $SKILL/resources/dailypings && python $SKILL/scripts/scrape_dailypings.py)
(cd $SKILL/resources/hn && python $SKILL/scripts/scrape_hn_200_days.py)
(cd $SKILL/resources/listyourtool && python $SKILL/scripts/scrape_listyourtool.py)
(cd $SKILL/resources/betalist && python $SKILL/scripts/analyze_betalist.py)
(cd $SKILL/resources/devhunt && python $SKILL/scripts/analyze_devhunt.py)
(cd $SKILL/resources/dailypings && python $SKILL/scripts/analyze_dailypings.py)
(cd $SKILL/resources/hn && python $SKILL/scripts/analyze_and_make_charts.py)
(cd $SKILL/resources/listyourtool && python $SKILL/scripts/analyze_listyourtool.py)
date +%Y-%m-%d > $SKILL/last-update.txt
Confirm when done:
Recalibration complete. Data refreshed across 5 platforms. Next recommended recalibration: {DATE + 30 days}.
