Build with AI · Logos Kit

These docs are written to be read by coding agents as well as people.


`/llms.txt`	An index of every page, one line each
`/llms-full.txt`	Every page's text in one file: paste it into your agent's context
Any page as Markdown	The Copy Markdown action on each page
`context7.json`	Lets Context7 index the repo for MCP clients
`AGENTS.md`	The repo's rules for agents: security, the QML sandbox, naming

A good first prompt:

Read https://logos-kit-docs.vercel.app/llms-full.txt. Then build a Basecamp
ui_qml app from `nix flake init -t github:logos-kit/logos-kit#dapp` that
connects a public account, shows its balance in LGO, and sends LGO with a receipt.
Handle user rejection, 6108 timeouts (onLateResult) and outcome "unknown".

Things agents get wrong in QML, and should be told:

amounts are strings: no Number(), no BigInt (use kit.sdk.parseUnits);
native amounts are lepta (1 LGO = 10^9 lepta): show kit.sdk.formatUnits(x, 9), parse input with kit.sdk.parseUnits(text, 9);
no Intl, no regex lookbehind or named groups, short is reserved;
no network access, remote or data: images, or Qt.openUrlExternally;
read kit.api fresh; it's rebuilt when the network changes.