Citius Holidays handles customer identity, travel, payment, and staff CRM data. Treat a suspected exposure as confidential and report it before discussing it in a public issue, social post, or customer channel.
The preferred intake is a private GitHub Security Advisory for this repository: open the repository's Security → Advisories → Report a vulnerability flow. If that flow is not available to you, ask the Citius director or engineering owner for the current confidential security contact. Do not send passwords, API keys, passport/visa documents, or full request cookies in a report; redact them and include only the minimum reproduction data needed to validate the issue.
Please include:
- affected route, component, or deployment and the first known version;
- a concise impact statement and realistic attack preconditions;
- reproduction steps or a minimal proof of concept that does not access other people's data;
- timestamps, correlation/request IDs, and redacted logs when available; and
- whether the report is already known to a customer or provider.
We will acknowledge a report within two business days, confirm severity and a mitigation owner as soon as the report is reproduced, and coordinate disclosure timing with the reporter. We may ask for a safe retest after a fix. Please do not exfiltrate or retain customer documents while proving an issue; stop after demonstrating authorization failure or the smallest harmless payload.
In scope are this repository, its deployed Citius web application, Convex functions and data access rules, authentication/session boundaries, CRM and account APIs, payment webhooks, and production configuration that is maintained by Citius. Third-party providers (Vercel, Convex, Resend, Razorpay, Sanity, and identity providers) are in scope only for a Citius integration mistake; report a provider-native vulnerability through that provider's own program as well.
Out of scope are denial-of-service testing, social engineering, physical attacks, spam, bulk account creation, and access to real customer documents beyond what is needed to prove a bug. Automated scanning must be limited to a test account and a low request rate. Never test against production payment capture or send messages to customers.
The release gate must retain the relevant test output, dependency audit, secret/policy scan,
deployment identifier, environment preflight, and WAF evidence for each production release.
Bootstrap Admin access is break-glass only and requires a future
PORTAL_BOOTSTRAP_ADMINS_EXPIRES_AT; staff provisioning records the actor and expiry in the
CRM activity log. See docs/SECURITY_WAF_RUNBOOK.md for the
operator baseline and incident evidence checklist.
If a public disclosure appears before a private report can be filed, do not include exploit details or sensitive data. Point maintainers to this policy and request a private channel.