The Kin ecosystem

Find your way around Kin.

Start with Kin to store and explore your code. Use the editor and agent connections when you need them. This page explains how the related projects fit together and which ones are available.

How this works in Kin

Kin is one system, not a shelf of separate products. Each repository below has one job in that system. Install one system, not a pile of unrelated tools.

Kin's semantic graph connects the engine, optional filesystem projection, editor, and bundled agent surface. Kin is open-source developer tooling. KinLab adds hosted collaboration, shared memory, and team controls. Kin is a public alpha, and open source under Apache-2.0. The repositories below are free to adopt on those terms, and Kin is built from them. KinLab is the hosted team product.

Open source Apache-2.0

Explore the engine, filesystem projection, retrieval, inference, editor, and agent tools in their public repositories on github.com/firelock-ai.

kinApache-2.0

Kin: the system of record

Kin is an open-source code repository for people and AI agents. It versions code and recorded relationships together. Give people a clearer basis for directing and reviewing AI-driven work.

kin-dbApache-2.0

kin-db: the semantic engine

An embedded graph engine, not a server you run.

kin-vfsApache-2.0

kin-vfs: transparent projection

Serve graph-backed files to ordinary filesystem tools.

kin-editorApache-2.0

kin-editor: the editor surface

Kin in your editor, not a fork.

kin-mcpApache-2.0

kin-mcp: the agent surface

Semantic context for any AI agent, over MCP.

kin-modelApache-2.0

kin-model: canonical types

The shared vocabulary of the graph.

kin-lspApache-2.0

kin-lsp: type-resolved enrichment

Type-resolved relationships for the graph.

kin-blobsApache-2.0

kin-blobs: content storage

Content-addressed storage for Kin.

kin-searchApache-2.0

kin-search: lexical retrieval

Lexical search under the graph.

kin-vectorApache-2.0

kin-vector: vector retrieval

Semantic vector search under the graph.

kin-inferApache-2.0

kin-infer: local inference

Pure-Rust embeddings, computed locally.

Private proof infrastructure external access pending

Internal supporting repositories are documented for architecture clarity, but they are not publicly readable or forkable today.

kin-benchPrivate support

kin-bench: proof harness

Pinned proof infrastructure. Public access is pending.

Hosted proprietary

KinLab adds shared repository graphs and memory, team controls, advisory AI review, and collaboration around open source Kin.

kinlabHosted

KinLab: hosted team product

Team collaboration built around open-source Kin.

Start locally. Add the team product when you need it.

Open a component above for its boundary and its source, or install Kin and run the quickstart against a repository of your own. The reasoning behind the license split is on the open source page.