The AI coding assistantthat learns how you work

Codistry adapts to you with every session, and your code stays on your machine.

Watch demo

Free during open beta · No credit card required

VS Code with the Codistry panel open: the agent traced an OAuth sign-out bug, cites the exact lines it read, has applied the fix to the editor, and waits for approval to run the tests

Codistry in VS Code, citing the code it read

Stop re-explaining your codebase

Codistry builds context from your repo and learns from every session. Ask for a change and it comes back written the way you would have written it.

the ask: "Add an endpoint to revoke API keys."

from your repo
  • key-service.tswhere key logic lives
  • result.tsResult<T>, ok(), err()
  • routes.test.tshow you test
from past sessions
  • “failable services return Result<T>”
  • “every change ships with a test”
codistry · after learning your repo
// key-service.ts
export async function revokeKey(
id: KeyId
): Promise<Result<void>> {
const key = await keys.findById(db, id);
if (!key) return err("not_found");
await keys.revoke(db, id);
return ok(undefined);
}
// routes.test.ts
it("revokes a key", async () => {
  • your service layer
  • your Result<T> error helper
  • your test conventions

the same ask without that context: logic inline in the route · errors swallowed as any · no test

It knows whose code this is

From your commit history alone, Codistry maps who owns each part of the codebase: your areas, the ones you share, and the ones that belong to teammates.

So it defers to you where you lead, writes in your conventions there, and explains more where you're new.

Codistry · Expertise
Codistry's Expertise view: the codebase tiled by area and colored by owner, with a plain-English read-out of the developer's strengths

the Expertise view, built from your commits · how expertise works

Private by architecture

Everything Codistry learns about your code lives in your editor and your repo. Cloud models see only what you choose to send.

your machine
VS Code
Your editor
Codistry
Runs locally
Your code
Indexed in place
air-gap ready· fully offline with local models
optional · encrypted
cloud APIs
GPT-5
When you choose
Claude
When you choose
Web search
When you choose
You choose what's sent

the whole system, end to end · what the extension sends

Under the hood

The learning runs on a complete agent: parallel sessions, sandboxed execution, and open integration points.

Context engine

Maps your repo into a knowledge graph: symbols, relationships, and how the code changes together over time. Kept current as you commit. Watch one build:

ready · watching commits98 symbols · 144 relationships · 6 areas
How it works →

Parallel sessions

Run several tasks side by side. Switching between them never cancels a run.

auth refactorrunning
fix flaky CI testneeds input
db migrationdone
See sessions →

Sandboxed skills

Specialized skills run inside isolated Docker containers, away from the rest of your system.

$skill: generate-pdf-report
→ isolated container · no host access
done · your system untouched
About the runtime →

MCP integrations

Hook up databases, APIs, and internal tools over the MCP protocol. Pre-built servers or your own.

PostgresGitHubFigmayour API
View integrations →

Bring your own models

Claude and GPT when you want them. Ollama or LM Studio when nothing should leave the building.

ClaudeGPTOllamaLM Studio
Supported models →

Three steps to your first session

No procurement, no pilot program.

  1. Install the extension

    Grab it from the VS Code marketplace, or email yourself an install link to open at your desk.

    Install page →

  2. Connect a model

    Paste an API key, or point it at a local model through Ollama or LM Studio.

    Supported models →

  3. Hand it a real task

    Open a repo and ask for an actual change, not a toy demo.

    Quick start guide →

No idea what to hand it?

By the time the index is built, Codistry has ideas of its own: work drawn from your recent commits and the areas you own. You won't start from a blank prompt.

Codistry · New session
Codistry's new-session panel suggesting three pieces of work drawn from the repo's recent commits, each labeled with the area it belongs to

drawn from your recent commits, refreshed as you work

Try it on your own codebase

Install the extension, open a project, and give Codistry something real to do. The first session is the demo.

Free during open beta · No credit card required

Prefer receipts to promises? What the extension sends·Privacy policy·FAQ