Skip to content

Security: Alex7develop/norma-ai-agent-platform

Security

SECURITY.md

Security Policy

Supported versions

The main branch is the actively maintained line for security fixes during the MVP phase.

Reporting a vulnerability

Please do not open a public GitHub issue for security problems.

Instead, email the maintainers using the address on the GitHub organization / profile, or open a private security advisory on GitHub if enabled for this repository.

Include:

  • Description of the issue and impact
  • Steps to reproduce
  • Affected component (API route, worker, auth, etc.)
  • Any suggested fix (optional)

We aim to acknowledge reports within a few business days.

Security baseline (current)

  • Secrets via environment variables (never commit .env)
  • HttpOnly cookie JWTs; refresh tokens hashed at rest
  • Workspace membership checks on tenant-scoped APIs
  • Knowledge/memory retrieval filtered by workspace_id / space_id
  • Worker and API process separation for long-running jobs

Production hardening still required

Managed secret storage, rate limiting, audit logging, TLS termination, and stricter tool/integration permissions before any public multi-tenant SaaS deployment. See docs/architecture.md.

There aren't any published security advisories