Daemon8 runs on developer machines and handles runtime output that may include sensitive application data. Security reports are taken seriously and prioritized above feature work.

Do not open a public GitHub issue. Send the report privately:

• Email: mail@daemon8.ai
• Subject prefix: [SECURITY] — helps triage

Include:

1. A description of the issue and its impact.
2. Repro steps or a proof-of-concept.
3. Affected component — the daemon binary (this repo) or a specific ingest endpoint.
4. Version or commit SHA of what you tested against.
5. Your disclosure timeline expectations, if any.

You will receive an acknowledgement within 72 hours. We aim to ship a fix and coordinated disclosure within 90 days. If a fix takes longer, we will tell you why and revise the target.

In scope for this policy:

• The daemon binary and its Rust crates (this repo).
• Ingest endpoints (HTTP /ingest, UDP when enabled, Unix socket when enabled).
• MCP handlers: connect/status/help, live-feed read/write/watch, browser actions, lens management, connections, and debug-session lifecycle tools.
• Chrome DevTools Protocol bridge — specifically, any path that could let a malicious page influence the daemon beyond its intended surface.
• Configuration file parsing and secret handling.

Handled separately (same reporting address):

• Marketing site — daemon8ai/daemon8-site.
• daemon8.ai production surfaces — report to the same email with subject [SECURITY] daemon8.ai production.

Out of scope:

• Social engineering, physical access, stolen-device scenarios.
• Denial of service that requires no vulnerability (resource exhaustion against a daemon running with default limits and no additional configuration hardening).
• Issues in third-party dependencies that do not affect Daemon8's actual usage of them — report those upstream and let us know so we can track.
• Vulnerabilities in software not shipped with Daemon8 that happens to run alongside it (your browser, your OS, your editor).

Good-faith security research is welcome. We commit to:

• Not pursuing legal action against researchers who report in good faith under this policy.
• Not filing a complaint with your employer, school, or hosting provider.
• Treating your report as confidential until a coordinated disclosure window closes.

In exchange, we ask:

• Do not exfiltrate data beyond what's necessary to demonstrate the issue.
• Do not publicly disclose until we've agreed on a timeline.
• Do not test against daemon8.ai production beyond passive analysis. If you need to probe live infrastructure, email first and we'll set up a staging target.

Post-fix, we publish a CVE (where applicable) and link the advisory in /CHANGELOG.md under the release that carries the fix. Reporters are credited unless they request anonymity.

Thanks for taking the time. If you find something, we want to know.