Commit Graph

2 Commits (265136fc42c49fe17273616332b4f297a24b098b)

Author SHA1 Message Date
boban 4d349eebb9 fix: alert mail on queue, SMTP reset, cert/uptime auto-probe, pill heights, XSS
Batch of user-reported fixes + two findings from the nav-split audit.

Bugs:
- Alerts never e-mailed. AppServiceProvider::applyMailSettings() ran only
  at boot; the long-lived QUEUE worker booted once, so SMTP configured
  after worker start (the normal case) meant queued alert/reset mail used
  the stale/log mailer. Re-apply the DB SMTP config before every job
  (Queue::before) so queued mail always sends with the current settings.
- SMTP could not be fully reset. Add Email::clearConfig() (+ a Reset
  button + Setting::forget) that removes every mail_* setting, reverting
  the panel to the safe default (log) mailer.
- New cert endpoint / uptime check showed no validity/status until a
  manual refresh. Certs\Index::addEndpoint + Health\Index::addCheck now
  probe the just-added row immediately.

Audit findings (096ba3c..HEAD):
- HIGH stored/DOM XSS: the per-server Terminal tab rendered the server
  name inside an Alpine x-text JS sink (`x-text="title || '{{ name }}'"`);
  Blade escaping does not protect that context. Move the fallback to a
  data-attribute read as a plain string (server-terminal + the host
  terminal for consistency).
- LOW: /terminal is now admin-only, but the command-palette `g t` leader
  key still sent every role there → 403. Drop it from the static map;
  render it as a role-filtered nav entry (admins only).

UI:
- status-pill is h-8 now (the small-button height) so a status pill and
  the action buttons in the same row line up (Services/Certs/Docker/…).

748 tests (new clearConfig test); R12-verified (uniform pill/button
heights). Marketing docs at ~/clusev-site reworded to remove the
contradictory domain/CLUSEV_DOMAIN install note.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-06 19:46:46 +02:00
boban 3cab72bf46 feat(security): TLS certificate expiry monitoring (feature 7/8)
Monitor the certificate expiry of configured TLS endpoints (the panel's own domain, fleet
services, anything reachable) and flag the ones due soon — before they lapse.

- CertService::check(host, port) reads the peer certificate NATIVELY via a PHP SSL stream +
  openssl_x509_parse — no SSH, no shell, so `host` is never interpolated into a command. Trust is
  intentionally NOT verified (verify_peer off): the goal is to report expiry even for a self-signed
  or already-expired cert. certInfo() (subject/issuer/expiry/daysLeft) is split out for unit testing
  against a generated cert; status() maps daysLeft → ok / warning (<30) / critical (<7) / expired.
- Certs\Index page (route /certs, `operate`-gated: route + mount + per-method). Add/remove endpoints
  (host validated as a hostname/IP — rejects scheme/path/shell chars; port 1-65535). Lazy per-endpoint
  scan, each guarded. Delete via signed ConfirmToken + R5 modal. Add/delete audited. Internal
  endpoints are allowed on purpose (monitoring internal service certs is legitimate; the check only
  reads a cert, it POSTs nothing).
- lang/{de,en}/certs.php + audit.php cert.endpoint_* + shell.nav_certs (de/en parity). No emoji.

10 new tests: service (certInfo extracts subject/expiry from a generated cert, status thresholds,
check errors on an unreachable endpoint), component (route gating, add + audit, reject scheme/
metachar host + out-of-range port, scan checks each endpoint, delete via confirmed token). 730 tests
green, Pint, lang parity, self-reviewed. LE issuance is a documented future (needs ACME on the host).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-05 20:57:53 +02:00