PERSONAL PRODUCTIVITY

Inbox triage agent: auto-draft easy replies, defer the rest to deep work

Watches new Gmail messages, classifies each as low-stakes or deep-work, drafts a reply for the easy ones.

CategoryPersonal Productivity
Enginepaperclip
Difficultyintermediate
Triggerevent
Steps5
Setup~15 min

How it runs

The automated pipeline, trigger to output.

  • TriggerNew Gmail message receivedGmailGmail
  • LogicClassify: low-stakes reply vs. deep-work itemOpenAI
  • ActionDraft concise reply as a Gmail draft (low-stakes)GmailGmail
  • ActionApply Focus Queue label and extract action summary (deep-work)GmailGmail
  • OutputPost queued deep-work items to Slack digestSlack

What it does

Every incoming email gets read once by an agent that decides whether it can be handled in seconds or needs real attention. Quick, low-stakes mail (confirmations, scheduling nudges, one-line acknowledgements) gets a ready-to-send draft created in Gmail. Anything that needs judgment is labeled and pushed into a focus queue, then summarized in a single Slack message so you context-switch on your own schedule.

When to use it

Use it when your inbox is a constant interruption and most of it is noise that still technically needs a reply. Ideal for operators, founders, and managers who want the trivial 60% drafted automatically and the important 40% batched for one deliberate review block instead of all-day pinging.

How it works

  1. 1A new Gmail message arrives and triggers the flow.
  2. 2The agent classifies it: low-stakes reply vs. deep-work item, based on sender, intent, and required effort.
  3. 3Low-stakes branch: it writes a concise reply and saves it as a Gmail draft for one-tap send.
  4. 4Deep-work branch: it applies a "Focus Queue" label and extracts a one-line action summary.
  5. 5A Slack digest message posts the queued items so you review them in one batch.

Set it up

What you configure once, before turning it on.

  1. 1
    Connect GmailRead, draft, send, label.
  2. 2
    Connect SlackChannels, DMs, threads, mentions.
  3. 3
    Connect OpenAIModels, embeddings, files.
  4. 4
    Set each agent's modelWe leave models unset so you pick the tier — fast + cheap, or top-quality.
  5. 5
    Tune it to your dataEdit the prompts, filters, and field mappings so it matches how your team works.
  6. 6
    Test, then turn it onRun once against a sample, confirm the output, then enable the trigger.

Run this workflow in your colony.

14-day trial. No DevOps. No Sales call. Provisioned in under a minute.