services/crypto.go keeps its function names and its KEY_ENCRYPTION_KEY
lookup and delegates the cipher, so vantagectl's verify probe can decrypt
with the same implementation rather than a second copy.
Thirteen tasks, each ending in a testable deliverable and a commit.
Also drops --log-level from the spec: the tool's whole output is what it is
telling the operator, and a level that could hide a key warning is worth not
having.
vantagectl/Dockerfile follows the repo's per-module convention rather than
living under deploy/docker.
verify's live check needs AES-GCM open, so the cipher primitives move to
shared/cryptobox and services/crypto.go delegates, rather than a second
copy of the cipher existing in another module.
Standalone vantagectl CLI (cobra, own module) that dumps and restores a
whole Vantage MongoDB database, stamping a sha256 fingerprint of
KEY_ENCRYPTION_KEY into the manifest so a restore cannot silently produce
a database whose secrets are unreadable.
The key itself never enters the archive.
- The status page URL was given only as `<instance>.vantage.<tld>`, which a
self-hosted install does not serve. Both deployments are now described.
- The banner is documented as one notice: the editor exposes no level picker
and the view renders every level identically.
- `pending` added to the component states, which a monitor with no result yet
renders.
- Delete page documented alongside un-publish.
- `TRUSTED_PROXIES` names the LAN case: with the RFC1918 default, a client on a
private range reaching the server directly is itself trusted and can spoof
`X-Forwarded-For` — and now `X-Forwarded-Host`. Narrow it to the proxy.
- CLAUDE.md: scopes are nine resources, not eight; `status-pages` added to the
REST route table; the host-resolution rules recorded under Status pages.
- The display-name placeholder showed the monitor's own name, reading as "leave
this blank and we will use it". The server deliberately does the opposite: a
blank `display_name` publishes the raw monitor id, because publishing an
internal name has to be a decision. The placeholder now says "Public name
(required)" and Save is refused until every component has one, so nobody adds
five monitors and discovers five UUIDs on their public page. The server
fallback is unchanged.
- `deleteStatusPage` and `deleteStatusIncident` existed in the api client and
were wired to nothing, and the page address is immutable — delete was the
only correction for a typo and there was no way to reach it. The editor
header gains a typed-confirmation Delete page, and each incident row a
confirmed delete, both on the existing ConfirmDialog.
- The create modal's address hint had lost its em dash and read as a broken
sentence.
- A derived monitor outage with no `resolved_at` went to `History`, so an
ongoing disruption was listed under "Past incidents" while the component pill
beside it read Down. Unresolved now goes to `ActiveIncidents`.
- `overallState` returned `up` when nothing was counted: "all systems
operational" claimed from no evidence at all. A page with no components now
reports `no_data`, which the view already renders as "Status unknown".
- `EnsureStatusPageIndexes` returned on the first failure, so a transient
failure on the `status_pages` index left `status_incidents` with no unique
`(instance_id, incident_id)` index — a correctness property, not a scan
optimisation. All three are attempted and the failures joined.
The SSR fetch set `Host` to the visitor's hostname. `Host` is a forbidden
header name and undici discards it silently, so the Go server saw
`server:8080`, `hostSlug` returned "", `InstanceFromHost` returned false and
every public status page 404'd on every deployment. The feature did not work.
- `web/` now forwards the visitor's host as `X-Forwarded-Host`, and their
address on `X-Forwarded-For` — without the latter gin sees a request from the
Next pod with no XFF and every visitor of every page shares one 120/min
bucket, tripped by exactly the traffic an outage produces.
- `publicStatusInstance` honours `X-Forwarded-Host` only when `c.RemoteIP()` is
in `TRUSTED_PROXIES`. It is a tenant selector, so an untrusted peer must not
be able to name one; `RemoteIP()` rather than `ClientIP()` because the latter
is reconstructed from the very headers being judged. `TrustedProxies()` moves
from main.go into the api package so the variable keeps one parser.
- A host naming no slug on a non-cloud deployment resolves the sole instance,
the way bootstrap does. A self-hosted install at vantage.acme.com or an IP
has no slug and could never serve a status page; more than one instance is a
404 rather than a guess, and an unknown-but-well-formed slug stays a 404.
- `InstanceFromHost` gains an explicit-host variant rather than a second copy
of the slug rules, and now caches negative lookups: an unknown host cost a
Mongo query per anonymous request, which is also a timing oracle separating
"no such instance" from "instance exists, page does not".
- The handler's `@Router` annotation is dropped. openapi.json declares one
server of `/api`, so it published `/api/public/status/{pageId}` — a path that
does not exist. The real address is described in prose instead.
Adds owner|admin routes under /api/status-pages for authoring status pages
and their incidents/maintenance windows, gated by the status_pages licence
feature. Adds the "status" token scope resource and the ten route-scope
entries, and regenerates the committed OpenAPI document.
Also types ErrPageInvalid as a sentinel for status page/incident validation
failures (previously bare errors), so statusPageError maps them to 400
instead of 500, and createStatusIncident/updateStatusIncident route through
the shared error mapper rather than hand-rolling a 400 for any service error.
/api-keys shares a raw string prefix with /api, and the proxies in front
of this app do not all match by path segment. Nginx Proxy Manager routes
/api straight to the Go server with a prefix location, so /api-keys never
reached Next at all — it reached a control plane with no such route and
came back as a JSON 404. Traefik's PathPrefix has the same shape of
matcher, which puts the Helm ingress at risk whenever ingress.api.enabled
is on.
The page is /tokens now, which cannot collide with anything, and which
matches the /api/tokens the REST API already publishes. The sidebar still
says API Keys — the label is for the reader, the path is for the router.
A permanent redirect covers anyone who bookmarked the old path today.
Fixing the proxy config instead would have left the trap set for the next
deployment, and for whatever sits in front of it.
The page was hand-rolling its loading spinner, error line and empty
paragraph while the rest of the console routes these through
AsyncBoundary with a TableSkeleton and an EmptyState — the same drift
Async.tsx was written to end. It also passed className="p-0" where Card
takes padding={false}.
The admin-only scope switch becomes the vulnerabilities page's pill
filter rather than a loose checkbox: seeing everyone's keys is a filter
over the list, not a preference, and someone who has learned one control
has now learned both.
Scopes collapse to one chip per resource with an r/rw qualifier. Sixteen
badges made the row taller than everything around it and still had to be
read one at a time.
In the create dialog Copy is now the primary action and Done the quiet
one, because the value is unrecoverable once the dialog closes, and the
result panel repeats the role, scopes and expiry that were just granted.
The token management card sat on /settings, which is owner|admin
throughout, so it hid a capability every member already had: the API has
never required a role to mint or revoke your own key. It is now the
/api-keys page, reachable at every role, with the instance-wide lifetime
cap left behind on /settings because that is policy rather than one
person's credentials — and that split is what lets the page be ungated.
The sidebar gains groups: Fleet, Access, Automation, Instance, each with a
small-caps heading and a rule above it. Grouping is by what the operator
is doing rather than by which service answers, so SSH keys, secrets and
API keys sit together as credentials. A group whose every item is
admin-only disappears whole for a member; a labelled section with nothing
under it reads as a failure rather than a restriction.
The UI says keys while the collection, prefix and routes still say tokens.
Renaming a published endpoint to match a nav label would break every
script already written against it.
handleInstallScript and handleUpdateScript are registered on the bare
gin engine at /install and /update, outside the /api group the
generated document's BasePath assumes. Their @Router annotations
therefore published /api/install and /api/update, paths that 404 —
the reference page told a reader to curl a URL that does not exist.
Removed the swag annotations from both handlers (replaced with a plain
comment explaining why) rather than adding a corrected @Router, since
swag has no per-route BasePath override and there is nothing lost by
leaving two shell-script endpoints out of a JSON API reference — their
.ps1 counterparts were already undocumented for the same reason.
Regenerated internal/api/docs/openapi.json accordingly.
Every request presenting an expired token wrote a token.expired_use
audit row, and RateLimitTokens only applies once a session exists, so a
rejected token was never rate-limited. A looping job with one expired
token could write an unbounded number of audit rows, drowning the real
audit trail.
services.ShouldLogExpiredTokenUse now dedupes to at most one
token.expired_use record per token per minute, mirroring the throttle
TouchAPIToken already uses for last-used. It lives in services rather
than auth because the storage concern belongs beside the token's other
storage-backed state. The first use per window is still recorded, which
is what makes a forgotten job visible.
CreateAPIToken capped a new token's role at the creator's role but never
capped its scopes against the calling credential's scopes, and POST
/api/tokens required only settings:write. A token holding settings:write
alone could therefore mint a token holding keys:write or secrets:write,
reaching every SSH private key and vault secret in the instance.
createToken now refuses (403 scope_confinement) when the calling
credential is itself a token and any requested scope is not satisfied by
that token's own scopes, via services.ScopeSatisfied so servers:write
still permits granting servers:read. Cookie sessions are unaffected,
since their authority is the user's role. Also correct the createToken
doc comment, which claimed the scope cap already existed.
Also document why Hint stores 5 hex characters of the token secret.
swag v2.0.0-rc5's parseSecAttributesV3 resolves a security scheme's map key
via getSecurityDefinitionKey(lines), which scans from the start of whatever
comment-line slice it was handed and returns the first @securitydefinitions
match — ignoring the current parse position entirely. Three
@securityDefinitions.apikey blocks stacked in one Go comment group (the
three were separated only by bare '//' lines, which do not split an
ast.CommentGroup) therefore all resolved to the first block's name
(cookieAuth), with the last block's in/name/description winning: the
generated document had exactly one securityScheme, keyed cookieAuth, body
esoAuth.
Separating the three blocks with real blank source lines splits them into
three distinct ast.CommentGroups, so swag's file-level comment scan (which
requires no other tokens between them, same rule Go uses for doc comments)
hands each block its own line slice and each resolves its own key.
Regenerated openapi.json now carries all three schemes with correct
bodies, referenced with no dangling security requirements.
Generated from swaggo v2 annotations, committed rather than built into the
image: the runtime stage is scratch and adding codegen puts the toolchain
in the build. CI regenerates and diffs, so an annotation edited without
regenerating fails the build — without that the annotations would drift
while still looking authoritative.
Scalar is vendored rather than loaded from a CDN, because air-gapped
self-hosted installs are supported and a reference page that fails closed
offline is a support ticket.
Same treatment: named types replace gin.H literals, and every handler gets
a swaggo doc block. This is the last of the handler files under
server/internal/api/.
Same treatment: named types replace gin.H literals, and every handler gets
a swaggo doc block. VulnSummaryResponse uses pointer fields so the
db-freshness block stays entirely absent when no vulndb_meta document
exists yet, matching the handler's original conditional gin.H exactly.
Converts their gin.H responses to the named types added in the previous
commit and adds swaggo doc blocks for every handler in handlers.go and
tokens.go.
Named response types for handlers that were returning anonymous gin.H
literals, so a generated annotation and what the handler actually returns
cannot disagree. main.go carries the top-level swaggo info block (title,
description, security schemes for cookie, bearer token and ESO auth).
A card in the Access group beside Members and single sign-on rather than a
new nav entry — /settings/instance was folded back in for exactly this
reason. The plaintext is shown once in a well block and never again.
Tokens outside a newly tightened lifetime policy are flagged rather than
broken, because the policy governs issuance, not existing credentials.
600 per minute per token, in the Redis that sessions already require.
Cookie sessions are untouched. A Redis failure falls through rather than
refusing traffic — it is already a larger problem and should not become a
second outage.
createToken's catch-all mapped every unmatched error to 400, so a
database outage reported itself as a malformed client request. Wrap the
genuine validation failures with ErrTokenInvalid and let the handler
answer 500 with a fixed message for everything else.
createToken's catch-all answered 400 for every unmatched error, so a
database failure reported itself as the caller's malformed request. Found
in review of Task 7.
Create, list and revoke, with no update: editing what a credential already
deployed in CI can do, with no record of what it could do before, is worse
than requiring a rotation. Revoking a token that is not yours answers
not-found, since a 403 confirms it exists.
The audit actor stays the human and names the credential alongside, so a
person clicking and their CI job are told apart.
Keyed on the registered gin route pattern rather than a per-route
decorator, because a route registered without a decorator would be
unguarded. An unmapped route reached by a token is a 403, and a boot-time
check refuses to start when any /api route is missing, so the failure
lands at deploy rather than as a customer's surprise 403.
sessionFromCookie already writes "session expired" when a cookie was
presented and rejected with no bearer to fall through to. Middleware
called sessionFromToken anyway, which wrote a second "not authenticated"
body onto the same response for every ordinary browser-session timeout -
gin logged "superfluous response.WriteHeader call" on ordinary use, not a
rare edge case.
Guard on c.IsAborted() after sessionFromCookie: true only in that one
rejected-cookie-no-bearer branch, so it short-circuits there while the
other three credential paths (no credential, bearer only, stale cookie
plus valid bearer) are unaffected.
The plan's Middleware called sessionFromToken even when sessionFromCookie
had already answered a rejected cookie, putting two JSON bodies on the
wire for the ordinary expired-session case. Found in review of Task 5.
One middleware, two ways to arrive at the same *Session, so every handler,
role guard, licence gate and audit call is untouched. The host guard
applies to both: a token carries an instance, and the tenant boundary must
not have a token-shaped hole in it.
The effective role is min(user, token) recomputed per request, so demoting
somebody demotes their tokens with them. A stale cookie beside a valid
bearer falls through rather than refusing a credential that would work.
Mint, resolve, list and revoke, with the effective role capped at the
owner's and recomputed per request rather than frozen at creation.
Deleting a user deletes their tokens in the same call, so offboarding is
one action. Revoking somebody else's token answers not-found rather than
forbidden, since a 403 confirms the credential exists.
Also re-exports shared.APITokenMaxDays into server/internal/models,
following the existing ValidRole wrapper pattern, since the token
service needs it and it was never re-exported.
A pointer with absent meaning no cap, so an upgrade allows never-expire
tokens exactly as before and an instance opts into the policy. It governs
issuance only: changing it never invalidates a token that already exists.
Eight resources with read and write, write implying read. Coarse on
purpose: a scope per endpoint is a table nobody maintains, and a route
added without an entry either breaks or is unguarded.
The unique index on token_hash is what makes authentication an indexed
lookup rather than a scan, so this builder is fatal on failure like
EnsureAuthIndexes rather than warning like the secrets one.
Registered in ScopedCollections so instance purge reaches it.
Twelve tasks from model through middleware, scope enforcement, endpoints,
web UI, generated OpenAPI and documentation. Verification is build plus
curl and UI checks rather than test cycles, matching a repository with no
Go test harness beyond shared/mail.
Adds the approved design for personal access tokens on the control plane
REST API, and for the generated OpenAPI 3.1 document served as a Scalar
reference page.
Tokens fall back into the existing session middleware rather than getting
their own route group, so every handler, role guard and audit call works
unchanged. Scope enforcement derives from the registered route pattern and
fails closed, with a boot-time check for unmapped routes.
A Terraform provider is deliberately left to a follow-on spec.
mrhid6