feat: activity feed — enriched audit events + global/project timelines (#45)

* feat(branding): white-label — public /branding, admin write, shell theming

app_settings branding.* keys (product name, logo/favicon URLs, accent
color, docs/support links). Public GET /branding so login and shell
theme pre-auth; admin writes via PUT /settings.

useBranding() applies document.title, accent CSS var and favicon;
BrandWordmark replaces hardcoded wordmarks; admin Platform settings
gain a Branding form. MCP containr_branding_get + settings_update.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* feat(invites): team invite links — admin create/revoke, public accept

Admins mint single-use invite links (sha256-hashed token at rest,
optional email binding, configurable TTL). Public endpoints let the
accept page validate the link and register — working even while
public signup is disabled.

- migration + sqlc queries/model for user_invites
- POST /admin/invites (returns one-time token+url), GET list, DELETE revoke
- GET /auth/invites/:token public validation, POST /auth/accept-invite
- CLI containr invites list|create|revoke; MCP invites tools
- /auth/accept-invite page (expired/used/invalid states, bound-email lock)
- Admin page Team invites section: create, copy link, revoke
- OpenAPI + generated types + CLI guide

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* feat(admin): impersonate user — 15-min token, audit-logged

POST /admin/users/:id/impersonate mints a short-lived bearer token
(impersonated_by claim) for support/debugging. Admin accounts cannot
be impersonated; every mint is audit-logged with actor + target.

- CLI: containr admin impersonate <user-id> (works via CONTAINR_TOKEN)
- MCP: containr_admin_impersonate
- Admin UI: per-user Impersonate action → token modal with usage hint
- OpenAPI + generated types

Verified live: token resolves to target, admin target 403, audit row.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* feat(git): GitHub App manifest self-provisioning + app-level webhook

Admins create the instance GitHub App in-browser instead of hand-
setting GITHUB_APP_* env vars:

- POST /admin/git/github-app/manifest builds the manifest + URL;
  the frontend auto-submits it to github.com
- /git/github-app/callback page exchanges the code via
  POST /admin/git/github-app/convert; credentials persist in
  app_settings (precedence over env)
- GET /admin/git/github-app reports configured/source
- POST /api/git/github-app/webhook — single app-level push receiver,
  HMAC-verified, dispatches via extracted dispatchPushToServices
- install-url + installation-token paths resolve settings→env
- CLI: containr admin github-app status|manifest|convert; MCP tools
- Git providers UI shows "Set up GitHub App" for admins when
  unconfigured

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* feat: activity feed — enriched audit events, global + per-project timelines

audit_logs gains severity/category/label columns derived from
resource+action at write time; legacy rows are classified on read and
backfilled at boot in batches.

- GET /activity: global feed scoped to the caller's visible projects
  (admins see all); severity/category/resource/project_id/since
  filters, page+limit
- GET /projects/:id/activity: timeline correlated to the project's
  services, deployments, databases and the project record; access
  mirrors project read rules
- Retention worker prunes rows older than AUDIT_RETENTION_DAYS
  (default 90, 0 disables), running once at boot then daily
- project.create/delete now audited so timelines aren't empty
- CLI `containr activity` with filter flags; MCP tools
  containr_activity_list + containr_project_activity
- Frontend: /activity page (severity chips, icon per category) under
  Operate nav; project workspace gains an Activity view tab
- OpenAPI + generated types + CLI docs

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

---------

Co-authored-by: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

app/backend/internal/api/activity_test.go

@@ -0,0 +1,32 @@
+package api
+
+import "testing"
+
+func TestClassifyAuditEvent(t *testing.T) {
+	cases := []struct {
+		resource, action          string
+		wantSev, wantCat, wantLab string
+	}{
+		{"service", "service.create", "success", "service", "Service created"},
+		{"service", "service.delete", "warning", "service", "Service deleted"},
+		{"deployment", "deploy", "success", "deployment", "Deployment deployed"},
+		{"project", "project.create", "success", "project", "Project created"},
+		{"user_invite", "user_invite.create", "success", "system", "User invite created"},
+		{"user_token", "user_token.revoke", "warning", "security", "User token revoked"},
+		{"service", "deploy.fail", "error", "service", "Service fail"},
+		{"user", "impersonate", "warning", "user", "User impersonated"},
+		{"webhook", "update", "info", "webhook", "Webhook updated"},
+	}
+	for _, tc := range cases {
+		sev, cat, lab := classifyAuditEvent(tc.resource, tc.action)
+		if sev != tc.wantSev {
+			t.Errorf("%s/%s: severity = %q, want %q", tc.resource, tc.action, sev, tc.wantSev)
+		}
+		if cat != tc.wantCat {
+			t.Errorf("%s/%s: category = %q, want %q", tc.resource, tc.action, cat, tc.wantCat)
+		}
+		if lab != tc.wantLab {
+			t.Errorf("%s/%s: label = %q, want %q", tc.resource, tc.action, lab, tc.wantLab)
+		}
+	}
+}