LaunchRepoThe six steps
Everything below happens on your machine, with your accounts. Requirements: Python 3.10+ on macOS or Linux, a coding agent that can read files and drive a browser, and a Chrome you are signed in to. Nothing else is installed.
Step 1: Clone the invite
After payment you verify your email, connect the GitHub account that should have access and accept the invitation to the private repository. Clone it and open the folder in your coding agent — Claude Code, Codex, Cursor or another agent that can read files and drive a browser.
git clone <invite URL> LaunchRepoStep 2: Paste the start prompt
START.md contains one prompt: “Read AGENTS.md, then follow ‘Start the dashboard for the owner’ …”. The agent creates a private workspace next to the checkout — your products and credentials never live in the delivery repository — starts the local dashboard and answers with a 127.0.0.1 link.
prompts/start.mdStep 3: Describe your product once
In the dashboard forms or in an interview led by the agent (ONBOARDING.md) you fill product.json: audience, problem solved, differentiator, real features, pricing, launch stage, genuine screenshots and logo. identity.json holds the public name, contact address and nickname; a copy workshop produces the copy variants. You also set the earliest date anything may go live. Unknowns stay empty — the agent never turns planned features into available ones.
product.jsonStep 4: Connect your logged-in Chrome
BROWSER-SETUP.md attaches the agent to a Chrome you can see — Chrome DevTools MCP on your signed-in browser, or an isolated profile via CDP as a fallback — and runs a visible smoke check. You sign in to your mailbox in that browser so verification emails work. What the agent may do (free registrations, final submissions, specific email actions, badges on your website) is recorded once per product in authorizations.json. Spending stays at zero.
BROWSER-SETUP.mdStep 5: Your agent works the playbooks
doctor.py checks the setup and separates missing pieces from permissions you declined. Then the agent picks 3–5 fitting platforms from the index and catalog, and for each one: searches for existing listings, verifies it is on the right site, prepares copy within the field limits, declines every paid upsell and submits only within your grant. CAPTCHAs, 2FA and unclear account states stop the run and are handed to you.
playbooks/ · PLAYBOOK-INDEX.mdStep 6: Outcomes are recorded, not assumed
Every result goes through record.py into submissions.csv with one of eleven states — “live” needs a public URL and your confirmation. SESSION.md and FOLLOW-UP.md keep the evidence and the next dates, so the next session resumes instead of starting over. The dashboard groups it all as needs-you, problems, in review, scheduled and live. Your next product reuses everything.
record.py · submissions.csv
What still needs you
LaunchRepo is deliberately not a background service. These moments are handed to you and recorded as human handoffs instead of being “solved” by the agent:
- CAPTCHAs and bot checks
- The agent stops, records prepared_needs_human and the dashboard lists the platform under “needs you”. You solve it in your own browser.
- 2FA and verification codes
- Codes arrive in your mailbox in the connected browser. Forwarding a code by hand is a supported fallback and is recorded as a human handoff, never as automated access.
- Paid routes
- If the only usable route costs money, the platform gets deferred_paid. No purchase is ever made; you decide later, with the note in front of you.
- Your public representation
- Before a maker bio, launch intro or “why I built this” reaches a listing, PR-BRIEFING.md shows you concrete examples to correct. Approved wording is kept in representation.md.
- Website badges
- Some directories require a badge on your site. The agent asks once per product whether it may add, commit and deploy one; a “no” is recorded and does not block the rest.
- Confirming “live”
- A thank-you screen is not a listing. Marking a platform live requires a public URL and your attestation that you reviewed it.
What the tracker states mean
record.py accepts exactly these states, and the dashboard shows the same ones. Each is tied to a minimum observation so a “done” never gets inflated:
| State | Minimum observation |
|---|---|
draft | Work prepared; the final submission is not confirmed. |
prepared_needs_human | A precise owner action or challenge (CAPTCHA, 2FA, unclear account) blocks prepared work. |
submitted_pending_review | The platform explicitly received the submission and moderation is still pending. |
queued | The platform confirms a queue position or waiting-list state. |
scheduled | The platform confirms a launch date on or after your earliest launch date; any outstanding approval stays in the note. |
live | Public listing verified with no pending label; requires the public URL and your confirmation. |
already_listed | An existing public listing was found and verified; same URL and attestation requirement as live. |
blocked | Eligibility, account capacity or a mandatory missing fact prevents progress. |
deferred_paid | The currently usable route requires payment; no purchase was made. |
not_a_fit | The product does not meet the audience or listing criteria. |
unavailable | The route or platform cannot currently be used; the observed problem is noted. |
Each platform’s result stays independent: a partner-distribution checkbox on one site does not count as a second submission, and “scheduled” can still take several weeks to actually post on free-tier directories.
The founder explains it in under three minutes
Aleks walks through what LaunchRepo does, who it is for and what it deliberately leaves to you. Same video as on the home page; English audio.
Questions about the flow
Does it submit automatically while I sleep?
No. Your agent runs in your session, in a browser you can see, and the repository runs nothing in the background. CAPTCHAs, 2FA codes and paid decisions stop the run and are handed to you. You start the next batch when you want.
Which agents can run it?
Any agent that can read files and drive a browser: Claude Code, Codex, Cursor or a browser agent. The playbooks and instructions are English Markdown, which is what coding agents handle best; your product copy can be in any language a directory accepts.
How long until a listing is live?
From minutes to several weeks. Free-tier directories queue submissions for review, and “scheduled” is not “live”. Your earliest launch date sets a floor; the tracker records queued, scheduled and pending review separately so you always know what has actually posted.
What if a directory only offers a paid route?
The agent records deferred_paid with a note and moves on. The playbooks default to the free path and never authorise a purchase; whether a featured spot is worth it stays your decision.
Do I have to work in the terminal?
No. python3 scripts/dashboard.py starts a local dashboard on 127.0.0.1 with products, states, handoffs, follow-ups, the catalog and the prompts to hand back to your agent. Its forms write the same files the agent reads; --demo shows a sample workspace before you set up your own.
One payment, every product you launch.
The launch price is €99.98; it rises after every 20 sales. The pricing page shows the live price and everything that is included.
See pricing