LaunchRepoThere are, in 2026, three practical ways to get a software product listed on a few dozen directories without doing it all by hand: hand the job to a coding agent that can drive a browser, hand it to a dedicated browser agent, or pay a service to do it. They are usually compared on price. Price is the least interesting difference. What each one needs from you, who controls the accounts, and what you are left with when it is over matter more, and they are rarely spelled out.
This is the spelled-out version. It is written by the people who sell the first option, so read it with that in mind; where a service is the better answer, the article says so.
Option 1: your coding agent
A coding agent — Claude Code, Codex, Cursor’s agent, or anything similar that can read a repository and call tools — is not a submission tool. It becomes one when you give it three things: a browser it can operate, the facts about your product in a form it can read, and instructions per directory that tell it what the form wants and when to stop.
What it needs. A cloned repository on your machine. A browser integration: the toolkit documents Chrome DevTools MCP against a Chrome you are signed in to and can watch. Python for the tracker and dashboard. Your product facts in one file, your public identity in another, and a decision about which personal facts may appear on listings. Then one prompt.
Cost model. You pay for the agent’s usage — tokens, billed by your provider — and, if you use LaunchRepo, one payment for the repository. The second product costs agent time, not a second licence. There is no per-directory price, because nobody is doing the work per directory except your agent.
Control. Complete, and visible. The agent works in your browser with your accounts. When a CAPTCHA, a two-factor prompt or a verification email appears, it stops and records prepared_needs_human; you solve it in the same window. It never pays: a directory whose only route costs money is recorded as deferred_paid, and the decision is yours. Every outcome lands in a local tracker with a date and a public URL where one exists.
What you own afterwards. The accounts, the listings, the tracker with the real states, the playbooks, and a workflow you can run again next quarter or for the next product. Nothing lives on a vendor’s server.
Where it falls short. It needs setup you have to do once — browser connection, permissions, a first onboarding session — and it needs you within reach during a batch for handoffs. If you have never run an agent with a browser tool, the first hour is learning, not launching. The agent pages describe the setup for Claude Code, Codex and Cursor individually.
Option 2: a browser agent
Operator-style browser agents — hosted or local — already do the clicking. Point one at a submission form, give it text, and it fills the form. This sounds like the same thing as option 1 and differs in one important way: a browser agent, by default, has no memory of your product across sites and no idea what “done” means on any given directory.
What it needs. The same product facts and identity, but handed over per session, because a hosted browser agent cannot read your repository. A per-directory brief: what the form asks, what counts as a confirmation, what to refuse. A way to record results, since the agent itself usually cannot run your tracker.
Cost model. Usage billing from the agent’s provider, sometimes a subscription for the agent product itself. Plus whatever gives it the briefs; the browser-agent page explains how the toolkit’s playbooks serve that role even when the agent cannot read files directly.
Control. High while you watch, low when you do not. A browser agent will cheerfully click “Upgrade” if the free form is hidden behind it, submit a form twice if the first confirmation was ambiguous, or write “launched” in your notes when it saw a thank-you page. None of that is malice; it is the absence of rules. The rules have to come from somewhere.
What you own afterwards. The accounts and listings, if you signed in yourself. The record only if you kept one. A hosted browser agent’s session history is not a submission tracker.
Where it shines. Directories with heavy, visual forms where a coding agent’s browser tool is clumsy, and one-off submissions where setting up a repository would be overkill. Where it disappoints: batches of thirty platforms with follow-ups over several weeks. That is a state-tracking problem, and a browser agent has no state.
Option 3: a done-for-you service
You pay, you fill in a form about your product, and a team — human, automated, or, as several vendors now put it, “AI agents and humans together” — submits it to a fixed number of directories and sends a report. This is the largest market segment and the most varied.
What it needs. From you: product facts, a logo, screenshots, and sometimes access to an email address. From them: a package choice. Some create accounts on your behalf with an address they control and hand over the credentials at the end; some use yours; some sign in with their own Google account on directories that offer nothing else. That is the detail to read before paying, and the vendor comparison quotes what each of the eight vendors we checked publishes about it.
Cost model. Usually a one-time fee per product or per website, tiered by directory count; sometimes a subscription with a monthly quota. The second product costs the same again. The prices we found ranged from tens to several hundred dollars per product on the day we checked; the comparison pages have the dated numbers.
Control. Low, by design; that is what you are paying for. You see a report at the end. Whether the listings were on the free route, whether a form was submitted twice, and whether “listed” meant live or pending review depends on the vendor’s reporting standard. Some publish live URLs and screenshots; some deliver a spreadsheet.
What you own afterwards. The listings that went live, and account credentials if the vendor hands them over. Not the process: the next product is a new order. Some services attach a guarantee about domain ratings or listing counts; the honest ones also state that the directories decide, not the vendor.
Where it is the right answer. You have one product, no interest in agents, and a budget. Then a service is rational, and the comparison pages exist so you can pick one on facts rather than on the loudest landing page. The alternatives hub sorts the vendors by model.
The differences that decide it
Put the three side by side on the questions that actually vary:
| Coding agent + toolkit | Browser agent | Done-for-you service | |
|---|---|---|---|
| Who does the work | Your agent, in your browser | The agent, in its browser or yours | The vendor’s team or automation |
| Who holds the accounts | You | You, if you signed in | Vendor or you, per package |
| Price scales with | Agent usage | Agent usage | Products |
| Second product | Same repository | Same setup, new briefs | New order |
| Handoffs (CAPTCHA, 2FA) | Recorded, yours to solve | Yours, if you notice | Vendor handles or skips |
| Free vs paid route | Free by default, never pays | Whatever it clicks | Package-defined |
| Record of what happened | Local tracker, 11 states | Whatever you kept | Vendor’s report |
| Afterwards | Playbooks, tracker, accounts | Accounts | Listings, report |
The “second product” row is the one that separates buyers. If you ship one product a year, the service column wins on effort and the price difference is small. If you ship several, or relaunch after a rename, the repository column wins on both, and the browser-agent column depends entirely on how disciplined you are about keeping records.
What LaunchRepo is, in this frame
LaunchRepo is option 1 packaged: playbooks for 346 platforms, a product template, a tracker with eleven states, and a local dashboard, in a private repository your own agent runs. It also works as the missing rulebook for option 2, which is why the browser-agent page exists. It is not option 3 and does not want to be. There is no team, no queue and no report; there is a repository and one price for unlimited products you own.
Which is a long way of saying: choose by what you want to own afterwards. If the answer is “nothing, just the listings”, pay a service. If the answer is “the process”, get the repository and give it to the agent you already have.
Sources: the LaunchRepo repository (START.md, BROWSER-SETUP.md, RUNBOOK.md) and the vendor pages checked for the comparison pages on 17 September 2026. No vendor was contacted for this article.