Skip to content

Security: Thatgfsj/neuroweave-cortex

Security

SECURITY.md

Security Policy

Supported Versions

Version Supported
1.0.x yes
< 1.0.0 no

Reporting a Vulnerability

NeuroWeave Cortex is an infrastructure-level memory runtime for AI agents. Vulnerabilities may include:

  • Injection: malicious text content causing unexpected behavior in embedding/search paths
  • Data leakage: unauthorized access to stored memories
  • Denial of service: graph explosion or memory exhaustion attacks

Do NOT open a public issue for security vulnerabilities.

Report vulnerabilities to the repository maintainer via GitHub's private vulnerability reporting:
https://github.com/Thatgfsj/neuroweave-cortex/security/advisories/new

Include:

  • Steps to reproduce
  • Affected components
  • Potential impact
  • Suggested fix (if available)

Response target: 48 hours for critical, 7 days for moderate.

Security Design Notes

NWC stores user conversation memories. Defense in depth:

  1. No network ingress by default — NWC runs in-process, not as a server
  2. No default LLM integration — atom_facts and compiler use template providers unless explicitly configured
  3. Graph size limitsmax_total_anchors, max_edges_per_node prevent unbounded growth
  4. Write gate — pre-storage quality filter blocks noise/junk entry
  5. MCP server is optional — not loaded unless explicitly enabled

There aren't any published security advisories