LaunchRepoDefinition
An AI coding agent is a language-model assistant that works inside a project: it reads files, edits them, runs shell commands and calls external tools. Claude Code and Codex run in the terminal; Cursor embeds the same loop in an editor. What makes it an agent rather than a chat is the loop — it decides the next step from what it observed.
Coding agents are not browser agents by default. They gain browser control through an integration such as an MCP server that exposes a Chrome instance as tools.
Why it matters for a directory launch
A directory launch is file reading, form filling and record keeping — exactly the shape of task a coding agent already does in your repository. Using the agent you already pay for means no new subscription, no upload of your product to a third party and a visible browser you supervise. It also means the agent has to be told the rules: no spending, no evasion of platform controls, honest states.
How LaunchRepo handles it
LaunchRepo is written for any coding agent that reads Markdown, runs Python 3.10+ scripts and can drive a browser through an MCP server. AGENTS.md carries the rules every returning agent must follow; the start prompt is one paragraph. Pages on this site cover the specifics for Claude Code, Codex and Cursor. Agent usage is billed by your provider and is not part of the licence.
See the term in the repository
LaunchRepo is 346 directory playbooks, a product context file and a submission tracker your own agent runs. Pay €99.98 once; use it for every product you own.
See pricing