Introduces a new diagnostic command that walks a registry of check
modules against a store user's Webex Calling configuration and
surfaces per-issue adaptive-card remediation for the fixable ones.
Checks (services/voiceDiag/checks/): dnd, callForwarding, callWaiting,
callIntercept, voicemail, hoteling, executiveAssistant,
outgoingPermission, phoneOnline. Remediations offered for DND,
forwarding, waiting, and intercept.
Uses the /v1/people/{id}/features/* admin surface (spark-admin:people_read
+ spark-admin:people_write scopes we already hold) — the earlier
telephony/config/people/*/callSettings/* path scheme returns 404 from
the Webex gateway and is not a live surface. Runner distinguishes
routing-404s ("URL moved") from "not applicable" 404s ("no calling
license") via the response body.
Arg parser accepts detail/detailed/--detail/--detailed and normalises
macOS smart-dashes so --detailed doesn't die when auto-correct
turns it into an em-dash.
Wires a VOICEDIAG_ACTIONS dispatcher in index.js mirroring the IGMP
branch, and registers /voicediag in commands/registry.js. 170 tests
pass (52 new: 39 check + 12 renderer + 5 arg-normalization).
Docs updated in .env.example, services/phoneService.js:467, and a new
services/voiceDiag/README.md that includes a "how to add a check"
recipe plus a note on the earlier wrong URL scheme.
Co-authored-by: Cursor <cursoragent@cursor.com>
55 lines
2 KiB
JavaScript
55 lines
2 KiB
JavaScript
// src/services/voiceDiag/checks/executiveAssistant.js
|
|
//
|
|
// Info-only check for the Executive / Executive Assistant feature.
|
|
// This isn't something a store line would normally be configured
|
|
// for; when the type is anything other than 'UNASSIGNED', calls can
|
|
// be filtered/screened through an assistant relationship in ways
|
|
// that look like "the store phone isn't behaving right". We surface
|
|
// the current type so an operator can decide whether to unassign it
|
|
// in Control Hub. No remediation button — this is intentional
|
|
// executive configuration, we don't auto-flip it.
|
|
//
|
|
// Endpoint: GET /v1/people/{personId}/features/executiveAssistant
|
|
// Scope: spark-admin:people_read
|
|
// Response shape: { type: 'UNASSIGNED' | 'EXECUTIVE' | 'EXECUTIVE_ASSISTANT' }
|
|
//
|
|
// NOTE: Webex has announced a path migration for this endpoint (see
|
|
// developer.webex.com changelog — /v1/telephony/config/people/{id}/executive
|
|
// is the new path). When that ships and starts returning 404 on the
|
|
// legacy path, swap the ENDPOINT constant. Do NOT eagerly switch
|
|
// before Cisco's cutover window closes — until then only the legacy
|
|
// path is live.
|
|
|
|
const ENDPOINT = (personId) =>
|
|
`people/${personId}/features/executiveAssistant`;
|
|
|
|
export const executiveAssistantCheck = {
|
|
id: 'executiveAssistant',
|
|
label: 'Executive / Assistant',
|
|
requires: ['personId'],
|
|
scope: 'spark-admin:people_read',
|
|
|
|
async run(ctx) {
|
|
const data = await ctx.webex.request('GET', ENDPOINT(ctx.personId));
|
|
const type = data?.type || 'UNASSIGNED';
|
|
|
|
if (type === 'UNASSIGNED') {
|
|
return {
|
|
status: 'ok',
|
|
message: 'No executive / assistant relationship (normal for a store line).',
|
|
details: { type },
|
|
remediation: null,
|
|
};
|
|
}
|
|
|
|
return {
|
|
status: 'warn',
|
|
message:
|
|
`This line is configured as **${type}** in the executive/assistant feature. ` +
|
|
`Verify this is intentional — otherwise calls may be routed through an ` +
|
|
`assistant relationship.`,
|
|
details: { type },
|
|
remediation: null,
|
|
};
|
|
},
|
|
};
|