Please do not open a public issue for a security problem.

Email moshiur@ticonsys.com with a description, reproduction steps, and the impact you have assessed. You can expect an acknowledgement within 72 hours and an assessment within seven days.

v0.1.0 ships only the audit and geo-artifacts skills, and neither reads a credential or calls a third-party API today — Config.secret() exists and is exercised by tests, but no shipped skill calls it yet. The config schema already reserves the shape for when that changes. The areas most worth your attention, now and as later skills land:

• Credential leakage — a secret appearing in a report, log, or committed file. Config will hold env: pointers only; a literal secret is rejected at schema validation.
• SSRF via configured URLs — the crawler follows URLs from config and sitemaps.
• Write-scope escalation — anything letting a non-publishing skill perform a network write. No publishing skill ships until v0.7; when one does, it will be gated by design: dry-run default, explicit human approval, and confirmation before any outward write.

Rate limits imposed by third-party APIs, and findings that require an attacker to already control the machine running OmniRank.