Skip to content

Security: peg/snare

SECURITY.md

Security Policy

Reporting a Vulnerability

If you find a security vulnerability in Snare, please do not open a public GitHub issue.

Report it privately by emailing: security@snare.sh

Include:

  • Description of the vulnerability
  • Steps to reproduce
  • Impact assessment
  • Any suggested fixes

We aim to acknowledge reports within 48 hours and provide a fix or mitigation within 14 days for critical issues.

Scope

In scope:

  • Snare CLI (internal/, cmd/)
  • Cloudflare Worker (worker/)
  • Install script (install.sh)
  • Snare.sh API endpoints

Out of scope:

  • Theoretical attacks that require physical access to ~/.snare/
  • The fact that ~/.snare/manifest.json reveals canary locations to local users (by design — see Known Limitations in ARCHITECTURE.md)
  • Social engineering attacks against users

Security Model

What Snare protects against: Detection of compromised AI agents attempting credential use.

What Snare does not protect against: An attacker with local filesystem access who can read ~/.snare/manifest.json. The manifest intentionally stores canary locations and content for teardown purposes. Protect the manifest with file integrity monitoring if this is a concern.

Privacy guarantee: The /c/{token} callback endpoint never reads, stores, or forwards request bodies. Only header-derived metadata is stored (IP, user agent, timestamp, ASN). This applies to the managed snare.sh worker. Self-hosted deployments are controlled by the operator.

For a team/security review checklist covering data flow, files touched, proof artifacts, SIEM routing, and cleanup, see docs/enterprise-evaluation.md.

Device secret: The CLI generates a 256-bit device secret at ~/.snare/config.json (0600 permissions). This secret authenticates API calls to snare.sh. If compromised, run snare rotate immediately.

Supported Versions

Version Supported
latest release ✅
older releases ❌

We maintain only the latest release. Please update before reporting issues.

There aren't any published security advisories