Skip to content

Security: onurceri/botla

Security

SECURITY.md

Security Policy

Supported versions

Botla is developed on a single line. Security fixes go to the main branch and are not backported to older commits or tags.

Version Supported
main Yes
Anything older No

Reporting a vulnerability

Please don't report security issues in public GitHub issues, discussions or pull requests.

Report them privately through GitHub Security Advisories: open the repository's Security tab and click Report a vulnerability. Please include:

  • the affected component (API endpoint, widget, dashboard, ingestion, etc.) and the commit you tested
  • steps to reproduce or a proof of concept
  • the impact as you understand it, such as data exposure, cross-tenant access or code execution

If you can't use GitHub Security Advisories, open a public issue that asks for a private contact channel, without any vulnerability details.

Scope

In scope:

  • the Go API (cmd/, internal/, pkg/): authentication and session handling, tenant isolation between organizations and workspaces, plan and quota enforcement, admin authorization
  • URL and file ingestion, including SSRF and parser issues
  • the embeddable widget and its public chat endpoints: domain allowlist and embed-token bypasses, XSS in rendered answers
  • the dashboard (frontend/)
  • prompt injection that leads to a concrete security impact, such as leaking another tenant's data or calling actions the bot owner didn't configure
  • the Docker, Compose and CI configuration in this repository

Out of scope:

  • a model giving a wrong or off-topic answer with no security impact (that's a quality issue; please open a normal issue)
  • vulnerabilities in third-party services (OpenAI, OpenRouter, Qdrant Cloud, Cloudflare) or in dependencies that are already publicly known, unless Botla uses them in an unsafe way
  • denial of service through traffic volume, and findings that need physical access or a compromised developer machine
  • missing hardening headers or best-practice suggestions without a demonstrated impact

What to expect

  • Acknowledgement within 3 working days.
  • Initial assessment (severity and whether it's in scope) within 10 working days.
  • Fix: critical and high-severity issues are prioritised and usually fixed within 30 days. You'll get updates while the work is in progress.
  • Disclosure: once a fix is on main, an advisory is published. Reporters are credited unless they'd rather not be.

Botla is maintained by one person, so these are good-faith targets rather than contractual guarantees. Please give a reasonable window for a fix before disclosing publicly.

There aren't any published security advisories