We actively maintain and issue security fixes for the following releases:

Please do not report security vulnerabilities through public GitHub issues.

Report security issues by emailing deyzho@me.com. You should receive a response within 48 hours. If you do not, please follow up to ensure we received your original message.

Please include the following in your report:

• Type of issue (e.g., authentication bypass, SSRF, injection, credential leak)
• Full path of the source file(s) related to the issue
• Location of the affected source code (tag/branch/commit or direct URL)
• Any special configuration required to reproduce the issue
• Step-by-step instructions to reproduce the issue
• Proof-of-concept or exploit code (if possible)
• Impact of the issue and how an attacker might exploit it

This information helps us triage your report more quickly.

• We will acknowledge receipt of your vulnerability report within 48 hours.
• We will provide a more detailed response within 7 days indicating next steps.
• We will keep you informed of our progress towards a fix and public disclosure.
• We will notify you when the issue is fixed and coordinate public disclosure timing with you.
• We follow a 90-day disclosure deadline: if a fix is not available within 90 days, we will disclose the vulnerability with any available mitigations.

The Axon Python SDK implements the following security controls:

• SSRF protection — all outbound HTTP calls in axon.security.assert_safe_url() validate URLs against a private-IP blocklist and enforce HTTPS
• Secret filtering — environment variables with _KEY, _SECRET, _TOKEN, _PASSWORD, _MNEMONIC, or _PRIVATE_KEY suffixes are stripped by _filter_env() before injection into bundles
• Response size caps — all provider clients enforce a 1 MiB (4 MiB for GPU workloads) response limit
• Input validation — processor IDs and deployment names validated against allowlists to prevent path traversal and control character injection
• Prototype pollution prevention — remote JSON payloads sanitised with key blocklisting before use
• DNS rebinding defence — hostnames resolved to IPs before opening connections, then re-validated
• Credential isolation — .env files generated by axon auth are written with chmod 600 (owner-read-only)
• No secret logging — all logging paths strip credential-looking values before output

In scope:

• axon Python package and all its submodules
• The provider integration layer and routing engine
• Authentication and credential management
• CLI commands (axon auth, axon deploy, etc.)

Out of scope:

• Third-party provider infrastructure (io.net, Akash, Acurast, Fluence, Koii, AWS, GCP, Azure, Cloudflare, Fly.io)
• User-written workload code deployed through Axon
• Vulnerabilities in dependencies that have already been publicly disclosed — please report these upstream

We accept reports in English.