Audit log pages
Browsing the audit trail in the UI: project vs organization scope, the action and actor filters, JSON/CSV export with date presets, and the entry detail with metadata, field-level changes, and old/new snapshots. For what gets audited, the HMAC chain, retention, and SIEM forwarding, see Audit & SIEM.
Open it from the sidebar: Operations → Audit log.
Scope switch
A Project / Organization segmented switch in the header selects the view:
- Project — entries scoped to the current project (the project selector decides which).
- Organization — org-wide entries: logins, user and role changes, settings, license, LDAP, and everything not tied to a single project.
The audit list
| Column | Content |
|---|---|
| Timestamp | Local time of the event. |
| Actor | The username, or a system/… marker for automatic runs. A background-job row whose user was later deleted keeps its raw user/<id> marker (shown monospace with the full value on hover) so the id is never lost. A column filter lets you type part of an actor name (any case). |
| Action | Colour-coded tag with a friendly label — e.g. Certificate Issued (green), Login Failed (red), Settings Updated (blue), Distribution Dry Run (orange). A request that failed (HTTP 400 or higher) is always red and says so — Distribution Executed — rate limited, Target Created — denied (the role or CSRF check refused it), or … — failed. A signed-in user's request refused before it reached its handler (role gate, rate limit, bad body) is still recorded under the route's own action name with the targeted resource id; entries written by older releases under a raw POST /api/v1/… route are shown with that name too. Only a route CertAutoPilot does not know falls back to the raw METHOD path. |
| Resource Type | Human label — Certificate, User, DNS Credential, Module, Target, Approval Request, … |
| Resource | The resource's current name (certificate, target, credential, …) with the id on hover; a resource that was deleted since falls back to its truncated id. |
| IP | Source IP of the actor. |
- The Action column has preset filters by prefix: Auth, Certificate, Distribution, Job, Settings, User, Groups and access. The Resource column filters by resource type.
- The filters are applied on the server: they search the whole log, not only the page you are looking at, and the page numbers count the matching entries. They are kept in the address, so a filtered list can be shared as a link.
- Export takes the same filters as the list, together with the date range.
- Pagination uses the default page size from Settings → General (UI preferences).
Clicking a row opens the entry detail.
Action categories
The UI maps well over a hundred action codes to readable labels. The broad families:
| Family | Examples |
|---|---|
| Auth & sessions | login, login failed, logout, token refresh (refused ones only — a successful refresh is not recorded), token reuse detected, OTP verified/failed, TOTP setup/disable, password changed |
| Certificates | issue/renew/reissue/revoke requested, issued, renewed, revoked, failed/dead/cancelled variants, policy updates, deleted, downloaded |
| Approvals | approval requested (per operation), approved, rejected, executed, retried, cancelled |
| Distribution | executed, dry run, rollback (and rolled back), linked/unlinked, module/credential/target/target-group/path-set/action-set CRUD |
| DNS & CA | DNS credential CRUD + tested, discovery DNS provider CRUD + tested, zone CRUD, ACME account CRUD + validated, MSCA connection CRUD + tested + templates synced, CA connection CRUD + tested + CSR generated (CA_CONNECTION_CREATED / UPDATED / DELETED / TESTED / CSR_GENERATED, credential fields redacted) |
| Discovery & domains | source CRUD, scan triggered, finding status changed, domain tracking CRUD, check triggered |
| Admin | user CRUD, roles updated, password reset, sign-in method changes, project/organization CRUD, general settings, sign-in settings and SSO providers, LDAP, syslog, OTP policy, license, variables |
| Meta | job retried/cancelled, bulk action created, audit chain verified |
Entry detail
Clicking a row opens /audit/<id> with a Back to Audit Logs button and:
- Summary table — entry ID (copyable), timestamp, action tag, actor name and actor ID, source IP, project ID (or an Org-wide tag), org ID, resource type, resource name (when the resource still exists — a project-scoped resource only when it belongs to the entry's own project) and resource ID, and the entry checksum (the HMAC-chain value for this entry — see HMAC chain).
- Metadata — a key/value table of event-specific context, when the action recorded any.
- Changes — for update actions, a field-level diff table: Path, From, To.
- Snapshots — collapsible Old Value / New Value panels with the full before/after JSON documents, when the action captured them.
Sensitive fields (credentials, keys) are stored encrypted or redacted server-side and never appear in snapshots.
Export
The header has a date range picker (presets: last 7 / 14 / 30 days; default last 30) and an Export dropdown with two formats:
- JSON — one entry per line, suitable for SIEM ingestion or offline chain re-verification.
- CSV — for spreadsheets and quick audits.
The export respects the current scope: in Project view it exports that project's entries, in Organization view the org-wide log. The download opens in a new tab and is served by the API using your session — so it is subject to the same role checks as viewing the log. Larger, scripted exports (legal hold, DR) are better done straight against the export endpoint (audit export).
Entries are retained for 30 days (retention). If you need history beyond that, set up syslog forwarding or scheduled exports before you need them — the UI can only show what's still in the database.
Chain verification
There is no chain-verification button in the web UI, and no CLI command for it. The audit log is HMAC-chained, and verification is an API call (owner role, rate-limited to one request per minute):
curl -X POST -b cookies.txt -H "X-CSRF-Token: <token>" \
'https://<host>/api/v1/audit-logs/verify?limit=1000'
Page through a large log with since=<next_since> from the previous response. See Verifying the chain. Each verification run is itself recorded in the log as an Audit Chain Verified entry, so you can confirm from the UI that your scheduled verification is actually happening.
See also
Video tutorial
Short, hands-on videos showing this feature working for real. English audio, English and Turkish subtitles. All video tutorials →