MemorySync
Getting Started

Zapier Memory Automation

Give every Zap a long-term memory: three creates, three searches — including a prompt-ready Recall Context block for ChatGPT/Claude steps — and the category’s first New Memory trigger, so workflows can react when a memory is stored anywhere.

What the app does

TypeOperationWhat it does
TriggerNew MemoryFires when a memory is stored for an end user — from any surface: your app, an agent, a voice call, or another Zap
SearchSearch MemoriesSemantic search — results fan out as line items
SearchRecall ContextOne prompt-ready context block to paste into a ChatGPT/Claude step
SearchFind MemoriesNewest-first listing for an end user
CreateAdd MemoryStore a fact — server-side extraction keeps only durable content
CreateAdd Conversation TurnVerbatim capture with duplicate-proof idempotency seeds
CreateDelete MemoriesExplicit deletion by ID
Mem0SupermemoryZepMemorySync
Zapier app exists△ invite-only via a support email✗ a Code-by-Zapier snippet with an admitted bug in their docs✗ none✓ full app
Any trigger✗ zero triggers✓ New Memory — first in the category
Prompt-ready recall✓ Recall Context
Quota handling✗ 60s polling loops that outlive Zapier’s step timeout✓ clear upgrade error; silent free-tier quota never fails a Zap
Retry safety✓ idempotency seeds on turns

Recipes

1. Trigger — Gmail: New Email
2. Search — MemorySync: Recall Context
End User ID: {{from_email}}
Topic: drafting a reply to their email
3. Action — ChatGPT: Conversation
System prompt: You are my email assistant.\n{{context}}
4. Action — Gmail: Create Draft

Every operation is scoped by End User ID — map it from the trigger (an email address, a CRM contact ID, a phone number). One user’s memories can never leak into another’s Zap output.

The New Memory trigger

Zapier polls the newest memories for the end user you watch and deduplicates by memory ID, so each memory fires exactly once. Memories arrive from any surface — a voice call captured by the ElevenLabs proxy, an agent turn from LangChain, a manual add from your app — which makes this the bridge from “the AI remembered something” to “the business system reacted”. Polling interval follows your Zapier plan (1–15 minutes).

Output fieldExample
idm_789
textRenewed the annual plan and asked about SSO
sourcechat
created_at2026-08-24T09:00:00Z

Errors that behave like Zapier expects

ScenarioWhat happens
Monthly quota reached (evaluation keys)A clear QuotaExceeded error: “upgrade the plan or use a production key” — not an endless retry loop on a task that can never succeed
Free-tier silent quotaAdd Memory reports stored: false, accepted: true; reads return empty results — the Zap keeps running
Bad metadata JSON / malformed IDsA friendly validation error raised before any network call
Wrong API keyFails at connection time — the credential test makes a real query
Retried step after a timeoutAdd Conversation Turn carries a deterministic idempotency seed — the replay reports already_exists instead of duplicating

Availability

The app is complete and fully tested (33 offline checks, including the same schema validation the zapier validate CLI performs). It is available by invite while the public listing works through Zapier’s review process — ask us for an invite link, or use the n8n node if you self-host your automation.

Supported versions

SurfaceRequiresVerified on
memorysync-zapier 1.0.0A Zapier account (any plan); an invite while the listing is under review33 CI checks driven through zapier-platform-core’s own appTester, fully offline: auth + connection label, header middleware, idempotency seeds, tenant fallback, quota messaging, silent-quota envelopes, trigger dedupe ordering, and full zapier-platform-schema validation

Where to go next

Was this page helpful?