Docs assistant

Searches these docs locally; a configured answer service can generate responses.

Ask about this page or the wider Carina docs. Try: “How do I roll back a patch?”

llms.txt

Skip to content

Markdown

Overview

Carina is the local-first runtime for coding agents — policy, hash-chained audit, and transactional rollback.

Carina is a local-first AI agent runtime. It sits where a coding agent leaves chat and starts acting on a real repository — under policy, with an audit trail you can verify, and patches you can roll back.

RuntimeAgentsAudit

Three promises — everything else hangs off these:

PromiseMeaning
Policy before effectFourteen capabilities, profiles, approvals — the model never bypasses the kernel
Hash-chained auditAppend-only events; carina audit verify detects tampering
Transactional rollbackFile mutations via patches with provenance, not silent overwrites

Full power for the agent. Full control for you. Authority stays on your machine.

  1. Install

    curl -fsSL https://carina.nebutra.com/install.sh | sh

    More options: Installation.

  2. Run a task

    carina daemon start && carina doctor
    cd /path/to/your/repo && carina run "Summarize this repo"
  3. Prove it was governed

    carina audit SESSION && carina audit verify SESSION

Full tour: Quickstart · Recipes: Common workflows.

  • Engineers who want agents on local repos with audit and rollback
  • Platform teams building an internal agent runner
  • Security-minded teams watching files, shells, and secrets
  • Builders embedding a runtime behind IDE, TUI, CI, or web UI

Not primarily: a finished editor chatbot, or a hosted cloud agent product.

Clients talk JSON-RPC; the kernel gates effects; Zig tools execute.
SurfaceRole
CLI / TUILocal interactive coding and ops
JSON-RPC + GatewayEmbed in your product
SDKsTS / Python / Go typed clients
WorkersDistributed execution with leases
MCPExternal tools, still policy-gated

Same engine: session + profile + audit wherever you attach.

CategoryOptimized forCarina
Editor assistantsIn-IDE chatRuntime governance behind any client
CLI coding agentsConversational terminal UXDurable sessions, audit, rollback, workers, embed
Cloud agent tasksHosted remote UXLocal-first; cloud is optional boundary
Homegrown stacksOne-off orchestrationReusable policy / audit / patch primitives
  • Full editor product
  • Hosted managed agent SaaS as the core
  • Replacing Git history or code review
  • Pretending alpha packaging is finished enterprise distribution
  • Product framing: docs/product.md · cloud boundary: docs/nebutra-cloud-boundary.md
  • Architecture map: docs/architecture.md · docs map: apps/docs/FEATURE_MAP.md
  • Live CLI: carina --help · CLI reference
GoalPage
Install on my machineInstallation
First end-to-end runQuickstart
Day-to-day recipesCommon workflows
TermsGlossary

Was this page helpful?

One vote per page is recorded during this session.