Aug 21, 2026 — Apollo API billing outage persists (15th consecutive day); guardrail breach now 32 calendar days
- Re-confirmed today via the Apollo MCP connector (3 calls, different endpoints): identical
{"message":"API temporarily unavailable. There is an issue with your payment."}.
- Impact: unable to run the health check, evaluate guardrails, source/enroll leads, or check replies. Enrollment skipped (0 sourced, 0 enrolled). No changes made to sequences, copy, or volume.
- Recommended next step (15th consecutive identical flag): resolve the Apollo billing hold in Apollo's billing settings — this blocks everything downstream. Once resolved, pause both sequences and scrub both contact lists before any new enrollment. Recommending directly to Rahil that this scheduled task be paused until billing is fixed, since daily logging alone has not prompted action.
Hard rules (the daily task enforces these)
- Max 20 new enrollments/day (25 on manual days) · hard cap 50 sends/day
- Verified emails only — never enroll
unavailable/guessed/unverified
- Circuit breaker: 7-day bounce >5% OR spam-block >2% → STOP enrollments, investigate, log incident here
- No volume increases before deliverability score >80 (earliest: mid-Aug 2026) — domain age score is 1/5 (young domain)
- 600s minimum between sends (account setting — keep)
Account health — [email protected] (ID 6a44e0953378c0000ce3cbb9)
| Date |
Deliverability score |
Notes |
| Jul 6–12 (baseline week) |
69.6 — inbox placement "unhealthy" |
Hard bounce 6%, spam-block 4%, open 12.8% — driven by unverified sends + spammy copy |
Improvement plan
- Reply volume is the fastest reputation repair → step-1 reply-bait design
- Consider adding 1–2 warmed secondary sending domains if score <80 by mid-Aug (strategy doc suggests Litemail inboxes ~$20/mo)
- Check SPF/DKIM/DMARC on shilikajain.com (one-time manual check in Apollo settings)
Incident log
Sep 7, 2026 — Apollo access worsens again: API key itself now rejected outright ("Invalid API key"); guardrail breach now 49 calendar days, access outage 32nd consecutive day
- Re-confirmed today via direct curl fallback against 4 endpoints (emailer_campaigns/{id} for both sequence IDs, auth/health, mixed_people/api_search): all data endpoints now return {"error":"Invalid API key. See https://docs.apollo.io/reference/authentication for how to authenticate."} — a new, more severe error than Sep 5's "account no longer active" message. auth/health also flipped from is_logged_in:true (every day through Sep 5) to {"healthy":true,"is_logged_in":false} today — the key/session is no longer authenticated at all, not just account-restricted. The Apollo MCP connector (plugin:sales:apollo) also remains unauthorized/not connected in this non-interactive session, so only the curl fallback was tried, consistent with recent days.
- Impact: unable to run the health check, evaluate guardrails, source/enroll leads, or check replies. Enrollment skipped (0 sourced, 0 enrolled). No changes made to sequences, copy, or volume.