launch-directories
Skill4 starsChanged 5 months agoMITView on GitHub
Launch a product across startup/launch directories (Product Hunt-style listings) by driving the real logged-in boseclaw Camoufox session — one tested, reusable flow script per directory, with proof screenshots + a live-status audit. Use when the user says "launch <product> on the directories", "submit to directories", "run the directory signups", "do the directory launch", or wants to add a directory, re-run a launch, or audit which listings went live.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- alvaldes/launch-skillstheir own version, first seen Nov 8, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- alvaldes/launch-skills★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository13 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- autocomp-tickSkill
- ceoSkill
- fetch-protectedSkill
- homelabSkill
- initiate-ceoSkill
- initiate-ctoSkill
- initiate-gtmSkill
- initiate-qaSkill
- next-moveSkill
- plan-big-execute-smallSkill
- plan-small-execute-bigSkill
- solve-captchaSkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 64
Test with --isolated until the proof screenshot shows a real confirmation, then audit.mjs.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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.
Files
launch-directories — submit a product to launch directories, with proof + audit
Everything lives in directories/. The model is: the loop (Claude) drives the captured
boseclaw Camoufox session through each directory's submit flow — no API-keyed daemon, no fresh
logins (boseclaw is already signed into Google + GitHub; the session accumulates each new site's
cookies so re-runs are "hot login"). Each directory is a script; each run leaves proof;
an audit re-checks that listings actually went live. This is our repeatable GTM muscle — we
run it for every product we ship (Principle 17, distribution-first).
Layout
engine.mjs— Camoufox driver + action helpersH(goto/wait/waitFor/fill/type/click/clickIf/ select/upload/press/url/text/has/shot) + proof capture + ledger. Pre-downloadskit.logo→kit.logoFile.flows/<directory>.mjs— one proven flow per directory:export const meta = {directory, home, submitUrl}andexport default async function launch({H, kit}) { … return {status, listingUrl, notes} }. Statuses:submitted | already-live | queued | blocked | failed. Readflows/peerpush.mjs(OAuth + AI-autofill + logo upload + free-queue + decline-upsell),flows/awesomeindie.mjs(OAuth + form + tag-select),flows/pitchwall.mjs(dedup + AI-generate) as the templates.kits/<product>.json— the reusable product data (name, url, tagline, summary, description, categories, pricing, logo, positioningRule). One per product; every flow reads from it.launch.mjs— runner.audit.mjs— live re-check.merge-rows.mjs— fold parallel--isolatedruns.proof/—submissions.jsonledger +<slug>-<directory>.pngproof shots +audit-*.png.
Verbs
Launch a product on one directory (or all):
cd directories
node launch.mjs peerpush --kit kits/weekly-brief.json # one directory
node launch.mjs --all --kit kits/weekly-brief.json # every flow, sequentiallyLoads the boseclaw session, runs the flow, saves the session back (hot login persists), writes a
proof screenshot + a ledger row. --headful to watch it; --state <file> for another identity.
Audit which listings are actually live (run daily / from a sub-agent):
node audit.mjs # re-check every ledger row, fresh screenshots
node audit.mjs --slug weekly-brief # just one productMarks each 🟢 live (listing loads + name present) / 🟡 pending (queue/review wording) /
🔴 missing (404 or name absent → needs resubmit). Rewrites the ledger's audit field + timestamps.
Note: directories with no public per-product URL until approval (e.g. awesomeindie, hot100) read as
missing while pending — eyeball those from the proof shot rather than trusting the flag.
Parallelize a big batch (one subagent per directory): each subagent authors its flows/<dir>.mjs
and runs node launch.mjs <dir> --kit … --isolated (writes proof/row-<slug>-<dir>.json, does NOT
touch the shared ledger/session → no races). Then merge: node merge-rows.mjs.
Adding a new directory (the "script per directory" step)
- Copy the closest template from
flows/(OAuth-form → awesomeindie; AI-autofill+logo → peerpush; dedup+generate → pitchwall). - Rewrite the selectors/steps for the new site. Develop by dropping
await H.shot('probe')steps and Reading the PNGs inproof/; dump form field selectors withnode ../browser/camoufox.mjs act steps.json --state ../.secrets/boseclaw-state.jsonusing a{"fields":true}step. - Prefer OAuth ("Continue with Google/GitHub" → click
text=Bose Claw→text=Continue): it usually skips the site's CAPTCHA entirely. Camoufox's real fingerprint passes Cloudflare Turnstile/interstitials passively; hCaptcha/reCAPTCHA image challenges may hard-block — then returnblockedwith proof, never fake a submission. - Test with
--isolateduntil the proof screenshot shows a real confirmation, thenaudit.mjs.
Rules (match the per-flow enforcement)
- FREE tier only — never enter payment, always decline featured/expedite/paid upsells.
- Positioning — obey
kit.positioningRule; for Weekly Brief: it RANKS newsletters, it does NOT summarize — never write "summarize". Overwrite any AI-generated copy withkit.summary/kit.description. - Never invent traction — leave metrics blank unless measured.
- Honest proof — a
blocked/failedrow with a screenshot is a real result; never record a submission that didn't happen (CLAUDE.md hard rules + principle 13). - Dedup — check for an existing listing first; return
already-liveinstead of duplicating.
Bot identity, not gated
Directory submissions under the boseclaw bot identity are pre-authorized GTM (CLAUDE.md) — run them, don't ask. The gate is only for spend, the human's personal identity, and email-domain reputation.
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
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.

