Skip to content

Security: FetchDoge/atomic-admission

SECURITY.md

Security Policy

What this repository is

atomic-admission bounds agent authority at the meter: it stops agent over-spend before bytes move by requiring vendor-side atomic admission checks with zero outbound authority.

Reporting a vulnerability

Preferred: GitHub private vulnerability reporting. Open the Security tab on this repository and choose Report a vulnerability. That channel is private between you and the maintainer, requires no email, and nothing is posted publicly.

If you cannot use that channel, open a minimal public issue stating only that you have a security report and how to reach you. Please do not include exploit details, proof-of-concept code, or affected-version specifics in a public issue.

Scope

In scope:

  • Any path where an admission token or spend authority can be bypassed, replayed, or forged.
  • Concurrency or race conditions allowing over-spend past configured limits.
  • Deserialization or validation failures in request processing.

Out of scope / stated plainly:

  • Deployment configurations outside the reference admission harness.
  • Network-level DoS attacks against the hosting server.

What to expect

Stage Commitment
Acknowledgement of your report within 7 days
Initial assessment and severity call within 14 days
Fix, or an agreed public disclosure coordinated with you

You will be credited in the fix or advisory unless you ask to remain anonymous.

What this policy does NOT offer

There is no bug bounty, and no monetary reward is offered or implied. This is an independent research project maintained by one person. What it can offer is a fast, honest response and public credit.

There aren't any published security advisories