Thanks to visit codestin.com
Credit goes to github.com

Skip to content

Security: unison-labs-ai/code-chunk

Security

SECURITY.md

Security Policy

Reporting a vulnerability

Please report security issues privately — do not open a public GitHub issue.

Email [email protected] with:

  • a description of the issue and its impact,
  • steps to reproduce (a proof-of-concept if you have one),
  • any suggested remediation.

We aim to acknowledge within 3 business days and to keep you updated as we investigate. We'll credit reporters who want it once a fix ships.

Scope

This repository is a client library (TypeScript SDK for code chunking and brain ingest). It holds no secrets and is not a security boundary — all authentication, authorization, workspace isolation, and rate limiting are enforced server-side by the Unison brain API. Reports about this library are most useful when they concern:

  • credential handling (how UNISON_TOKEN is read and transmitted),
  • dependency or supply-chain risks,
  • path traversal or injection in chunk path generation.

Server-side or account issues should also go to the same email address.

Credential handling

The library reads UNISON_TOKEN from the environment and sends it as Authorization: Bearer <token> only to the configured API host (UNISON_API_URL, defaulting to https://brain.unisonlabs.ai). The token is never logged, written to disk by this library, or transmitted anywhere else.

Never commit a real usk_live_... token to this repository. Use .env (which is .gitignored) and copy .env.example as a starting point.

There aren't any published security advisories