Thanks for helping keep ronin and its users safe.

Only the latest minor release line receives security updates.

Please do not open public GitHub issues for security problems.

Email a private report to: rohithkandula937@gmail.com with [SECURITY] in the subject. Include, where possible:

• A description of the issue and its impact
• Steps to reproduce (PoC, sample input, command line, environment)
• The ronin version (ronin version) and Python version
• Any suggested fix or mitigation

You'll get an acknowledgement within 3 business days and a more detailed response within 7 business days.

We follow a 90-day coordinated disclosure policy:

1. Day 0 — report received, triage begins
2. Day 0–7 — confirmation and severity assessment
3. Day 7–80 — fix developed, tested, and released
4. Day 90 — public disclosure, unless an extension is mutually agreed

Reporters are credited in release notes (let us know your preferred name/handle, or if you'd rather stay anonymous).

• Agent execution — sandbox escapes, command injection, or unintended code-execution paths in the core agent loop (ronin code, ronin chat)
• Tool guardrails — bypasses of the tool allowlist, approval gates, or the filesystem sandbox
• Secret handling — leakage of API keys, tokens, or environment variables via logs, sessions, memory files, or error output
• The secret scanner / write guard — inputs that should be blocked but aren't

• Third-party MCP servers and the services they connect to
• User-installed plugins and unofficial forks
• Issues requiring a pre-compromised host or attacker-controlled local config
• Issues only reproducible on unsupported versions
• Social engineering and DoS against infrastructure not operated by this project

We will not pursue legal action against researchers who make a good-faith effort to follow this policy, avoid privacy violations and service disruption, and give us a reasonable window to remediate before disclosing.