Torqon MCP Memory Loop
Store a memory in Torqon, retrieve the right context, then optimize it for an Agent reply inside Heym.
Shared by Maan ยท 2026-07-03
The full canvas, before you import it
Click any node to see its config.
Click a node to select it โ same as the Heym editor; the panel shows its settings.
3 nodes ยท Community-contributed
Torqon MCP Memory Loop
Use Torqon as a persistent memory and context optimization layer inside a Heym Agent workflow. This community template keeps the flow explicit: first store a memory, then retrieve relevant context, then optimize the context window before returning a response.
What this workflow does
- memoryRequest accepts a memory note, a question, and optional extra context
- torqonMemoryAgent starts the Torqon MCP server locally with
npx - The Agent calls
store_memorywith the memory note - The Agent calls
retrieve_contextwith the question - The Agent calls
optimize_contexton the retrieved context plus the extra context - memorySummary returns the stored-memory summary, retrieved context, and optimized context
Setup
Create a Torqon API key from dashboard.torqon.dev, then import this template into Heym. Open torqonMemoryAgent, find the Torqon MCP connection, and replace <user-api-key> in the TORQON_API_KEY env field with your own key. Heym runs the MCP server with npx -y @torqon/mcp@latest.
Connect an LLM credential to torqonMemoryAgent, run the workflow once with the sample fields, then swap in your own memory notes and questions.
Memory scope
Personal memory is tied to the API key, so only that key can read and write the stored memory. Workspace memory is shared across an organization, so member keys read and write to the same memory pool. The Torqon tools stay the same; scoping happens automatically from the key.
Notes
This first version intentionally uses the explicit store_memory -> retrieve_context -> optimize_context sequence instead of auto_process. The two-step memory loop is easier to inspect, and adding optimization shows where heavy context windows can shrink before the final Agent response.
Notes
This is a community-contributed template, shared by a member of the Heym community. Review the workflow and update any credentials or model settings before running it in your own workspace.
Curator note: Torqon shared the stdio MCP setup for this community template.
How to import this template
- 1Click Import โ Copy JSON on this page.
- 2Open your Heym and navigate to a workflow canvas.
- 3Press Cmd+V / Ctrl+V โ nodes appear instantly.
Discover more automations
- Personal ProductivityGmail Support TriageClassify each new email in your Gmail Primary tab, alert Slack about urgent ones, and mark every handled email as read.
- Personal ProductivityOutlook Flagged Email Follow-upsFlag an email in Outlook or Microsoft 365 and get a ready-made follow-up task in Slack, with a summary, a next step and a link back to the email.
- Personal ProductivityTeams Meeting Scheduler AgentAsk in plain words for a meeting with a colleague. An agent reads both Outlook calendars, picks a time you are both free and sends the Teams invitation.
- Personal ProductivityAvailable Time SlotsRead your Google Calendar availability from today through month end, with optional dates for a shorter window or up to 60 days.
- Personal ProductivityBook a Calendar with MeAccept a visitor email and selected time, recheck your calendar, then create a Google Meet event and email the invitation from your connected account.
- Sales & CRMSales Call MemoryTurn call notes into an account update and a memory graph of contacts, deals and objections.
- Marketing & SEOCampaign MemoryTurn campaign briefs into a status update and a memory graph of claims, assets and owners.
- Dev & IT OpsHTML Status PageCheck an upstream service and answer a browser GET with a rendered status page instead of JSON.
- Dev & IT OpsWorkflow Change Audit LogCapture every workflow create, update, and delete on your Heym instance and post a batched summary to Slack.
- Customer SupportHITL Support Reply AgentDraft a customer-facing support response, pause for human approval, then continue with the reviewed reply.
- Dev & IT OpsCodex PR Fix AgentSend a coding task to Codex, open a draft PR when it succeeds, and notify Slack when Codex needs follow-up input.
- Dev & IT OpsOpenCode Go PR Fix AgentSend a coding task to OpenCode Go, open a review-ready pull request, and return the implementation result.