Mitra availability: read paths respect require_mitra_ping=false

When the operator sets require_mitra_ping=false, the auto-offline sweep
early-returns (by design — "don't gate online status on heartbeat
freshness"). The three Valkey read paths still gated on heartbeat
freshness anyway, which trapped the system: sweep won't remove the
mitra from mitras:online, but readers reject them as stale. The customer
CTA stayed permanently disabled with no recovery.

Fix all three to skip the heartbeat-freshness check when require_ping
is off, matching the sweep's contract:
- computeAvailabilityFromValkey (customer beacon)
- isMitraReachable (extension service)
- findAvailableMitrasFromValkey (pairing candidate finder)

The Postgres fallbacks already did the right thing (is_online only,
no heartbeat compare); this aligns the Valkey hot path.

Also: PATCH /internal/config/mitra-ping now publishes config:invalidate
for require_mitra_ping and mitra_stale_after_seconds, and the subscriber
in mitra-status.service was widened to listen for both. Flipping the
toggle in CC now busts the 10s availability snapshot immediately instead
of waiting out the TTL.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

backend/test/services/mitra-status.valkey-mirror.test.js

@@ -224,6 +224,27 @@ describe('mitra-status valkey mirror', () => {
 
       expect(await isMitraReachable(m.id)).toBe(false)
     })
+
+    // Mirrors the autoOfflineStaleMitras "no-op when require_ping=false"
+    // contract: read paths must not gate on heartbeat when sweep doesn't.
+    it('returns true with stale heartbeat when require_ping=false', async () => {
+      const sql = db()
+      try {
+        await sql`
+          UPDATE app_config SET value=${sql.json({ value: false })}
+          WHERE key='require_mitra_ping'
+        `
+        const m = await createMitra({ callName: 'NoPing', isOnline: true })
+        await v().set(vkHeartbeatKey(m.id), new Date(Date.now() - 3_600_000).toISOString())
+
+        expect(await isMitraReachable(m.id)).toBe(true)
+      } finally {
+        await sql`
+          UPDATE app_config SET value=${sql.json({ value: true })}
+          WHERE key='require_mitra_ping'
+        `
+      }
+    })
   })
 
   // ---------- recomputeCapacity ----------
@@ -331,6 +352,35 @@ describe('mitra-status valkey mirror', () => {
       await invalidateAvailabilityCache()
       expect(await v().get('availability:snapshot')).toBeNull()
     })
+
+    // Regression: when an operator turns off the ping requirement, the
+    // auto-offline sweep is also disabled, so heartbeats may legitimately
+    // become arbitrarily old. The beacon must NOT filter those out — that
+    // would put the CTA in a permanently-disabled state with no recovery
+    // path (sweep won't remove the mitra; cache always re-computes false).
+    it('includes mitras with stale heartbeats when require_ping=false', async () => {
+      const sql = db()
+      try {
+        await sql`
+          UPDATE app_config SET value=${sql.json({ value: false })}
+          WHERE key='require_mitra_ping'
+        `
+        const m = await createMitra({ callName: 'NoPingRequired', isOnline: true })
+        // Heartbeat 1 hour old — well past any reasonable stale_after_seconds.
+        await v().set(vkHeartbeatKey(m.id), new Date(Date.now() - 3_600_000).toISOString())
+        await v().del('availability:snapshot')
+
+        const result = await countAvailableMitrasFromCache()
+        expect(result.available).toBe(true)
+        expect(result.count).toBe(1)
+      } finally {
+        await sql`
+          UPDATE app_config SET value=${sql.json({ value: true })}
+          WHERE key='require_mitra_ping'
+        `
+        await v().del('availability:snapshot')
+      }
+    })
   })
 
   // ---------- autoOfflineStaleMitras ----------