Questa pagina ti porta dall’installazione al primo fix verificato. La documentazione completa (in inglese) — comandi CLI, server MCP, GitHub Action, configurazione e concetti chiave — vive su docs.codna.ai.
Aggiungi Codna a Cursor o Claude in una riga. Il tuo agente può chiedere a Codna mappe di simboli, dipendenze, impatto sui test e bundle di evidenze mirati.
codna mcp start --repo .
GitHub App nativa
Installa la GitHub App per permettere a Codna di analizzare le issue, commentare con evidenze della causa radice e aprire pull request di correzione.
Codna si addestra sul mio codice? No. Codna è progettato per non addestrarsi sul codice dei clienti.
Posso fare self-host? Sì. Il self-hosting è disponibile per sempre, con opzioni on-prem e air-gapped supportate per Enterprise.
Frequently asked
Install with a single command from your terminal. The codna cli runs locally, so you bring your own API key and your code never leaves your machine unless you choose otherwise.
The MCP server for coding integrates natively with Cursor and Claude. Add it to your editor's MCP config and Codna's graph engine becomes available to your AI assistant without any extra steps.
The GitHub App monitors your repository and opens pull requests with verified fixes automatically. Every fix is validated by your own test suite before the PR is created.
No. Codna is self-hostable, and egress is fail-closed by default. You supply your own API key, and the system is designed so your code is never used for model training.
The deterministic graph engine maps the relevant code first, producing an evidence bundle measured at around 600 tokens. Head-to-head against Cursor, Codna used 5× fewer tokens and ran 1.7× faster, every fix test-verified (100%).
Codna supports 250+ languages, and has mapped 130 repositories in 9.2 seconds for zero LLM tokens. If your project compiles or resolves dependencies, Codna can graph it.