Skip to content

Security: madeinplutofabio/neural-computation-protocol

Security

SECURITY.md

Security Policy

Supported versions

This project is early-stage and evolving quickly.

Security fixes land on:

  • the default branch (main)
  • the latest tagged release (when applicable)

If you are pinning an older tag, you may need to cherry-pick fixes.

Reporting a vulnerability

Please do not open a public GitHub issue for security-sensitive reports.

Instead, use one of these channels:

  1. GitHub Security Advisories (preferred)
    Go to the repository’s “Security” tab and choose “Report a vulnerability”.

  2. If you cannot use GitHub advisories, open a minimal issue titled “Security contact requested”.
    Do not include details. A maintainer will follow up privately.

What to include

To help reproduce and fix quickly, include:

  • A clear description of the issue and impact
  • Steps to reproduce (PoC if possible, but keep it minimal)
  • Affected versions / commits
  • Any relevant environment details (OS, CPU arch, Rust/Node versions)
  • Suggested mitigation (if you have one)

Coordinated disclosure

If you report a vulnerability, we’ll aim to:

  • acknowledge receipt
  • reproduce and assess severity
  • coordinate a fix and disclosure timeline

We may request that you keep the report private until a fix is available.

Security assumptions (current scope)

NCP’s reference runtime is designed around:

  • WASM sandboxing (no filesystem/network for bricks by default)
  • explicit resource limits (time, memory, output size)
  • deterministic routing and structured results

However, integrations (Phase 3) will widen the threat surface. Treat all adapters and deployment environments as part of your security boundary, and run the benchmark + conformance suites in CI.

Non-security bugs

For non-sensitive issues (crashes, incorrect routing, schema validation errors), please open a normal GitHub issue.

There aren't any published security advisories