Docs menu

Documentation

Run Your First AI Partner Workflow with One Recurring Task

Choose one small weekly task, let an AI partner prepare the work for your review, and complete three reliable rounds before adding automation.

Updated Doc version 1.4

Give one recurring task to one AI partner, then check whether fresh input still produces a result you can review item by item. This guide uses weekly feedback triage, keeps human review in place, and uses three rounds to decide what to automate next.

A solo operator reviews a recurring input, an AI partner's prepared draft package, and an approved output tray
The smallest repeatable loop: stable input, AI preparation, human confirmation, then delivery.

Before you start

  • Prepare three batches of similar material. You should already know the correct result for the first one. Remove personal information the task does not need.
  • Choose a conversation for receiving results and set aside time to review them.
  • Create a partner in step 3 if you do not have one, or reuse a partner with a suitable role. For installation and the full creation flow, open Quickstart: Create Your First AI Partner and Verify a Task in the Docs navigation.

1. Choose a recurring task small enough to learn from

Look at the last two weeks and pick one task that meets all four conditions:

  • It appears at least twice a week and has a reasonably stable input source.
  • The AI partner can produce a draft, checklist, classification, or options first.
  • You can judge the result in a few minutes.
  • A mistake can be sent back for revision instead of paying, deleting, broadcasting, or shipping.
Good starters: feedback triage, meeting notes to follow-ups, research organization, or FAQ reply drafts. Wait on automatic payments, deletion, or public publishing.

2. Write a workflow card

Do not begin with a long prompt. Define these five parts first:

  1. Trigger: when the work starts, such as Friday afternoon or when a new feedback batch arrives.
  2. Input: where the material comes from: a folder, conversation, or meeting note.
  3. AI preparation: a reviewable output, such as three feedback groups plus open questions.
  4. Your confirmation: decisions that stay with you: priority, promises, price, and whether to send.
  5. Delivery: where approved work goes next: a todo list, document draft, or next-input tray.
Example: weekly feedback arrives → the AI partner merges duplicates and groups bugs, requests, and questions → I confirm priority and reply boundaries → approved items enter next week's todos and reply drafts.

3. Choose or manually create the partner responsible

  1. Open AI Partners in the sidebar and check for a partner with a suitable role. Reuse it if one is ready. Otherwise, select New AI Partner at the top right and choose Create Manually. If ready-made recommendations appear first, switch to configuring your own partner.
  2. Choose a ready platform as prompted, enter a name and role, and select an available model. For example, name it Feedback Organizer and ask it to group feedback, retain the original evidence, and flag uncertain items for review. Select Next, then choose only the skills this task needs.
  3. Review the details and select Create, or Create & Connect Channel to set up a conversation immediately. After the creation message appears, return to AI Partners and check the name, role, model, and runtime state.
  4. Open that partner's channel controls, finish any setup or confirmation, and send a short test message. Wait for its actual reply in the intended conversation. Then send the workflow card's input, output format, and review point, using the first batch with a known result to calibrate the task.
New AI Partner at the top right of AI Partners, with roles, states, models, and channel controls on partner cards
Check whether an existing partner fits the task, or use New AI Partner at the top right.

Check the responsible partner's role, state, and channel controls in AI Partners. Once its card appears, verify the actual feedback-triage result in the chosen conversation.

4. Run the first round manually and check four things

  • Complete: did it miss anything important in the raw material?
  • Traceable: can each conclusion point back to a source instead of filling gaps?
  • Decision-ready: is the result shaped so you can compare, edit, or approve it?
  • Within bounds: did actions that need your judgment stop at the review point?

Turn your corrections into a fixed checklist. “Remember next time” inside one conversation is not a repeatable rule.

5. Use fresh input for two more rounds

Use the checklist to correct format and omissions in round two. In round three, switch to fresh real material so you know the workflow did not merely memorize the first sample. After all three rounds can be checked quickly, add only one thing: a fixed trigger, automatic input gathering, or delivery after approval.

Expand one box at a time. If the new action fails, you should be able to return to the previous working loop.

Completion check

Keep the same workflow card and review checklist. Complete three rounds with results that match their inputs, flag missing or uncertain material, and stop for your decision where required. Record corrections from each round before adding one automated step. Three successful rounds check this task; they do not replace ongoing review.

Troubleshooting

  • Where do I create a partner? Select New AI Partner from AI Partners and choose Create Manually.
  • The partner is listed but no result arrives. Finish pending channel setup or confirmation and receive a test reply in the intended conversation before assigning the task.
  • The format changes every time: provide an explicit output shape and say that missing values stay blank.
  • It invents missing facts: require a source for each conclusion and move unsupported items into “needs review.”
  • Review takes longer than doing it yourself: shrink the task to one classification, one draft, or one checklist.
  • Results look good but still feel risky: keep human confirmation, then add run history and a clear rollback path.

Data boundary

Work data stays on your device by default; only channels you connect and model calls go online as the task requires. Connect only what this workflow needs and provide only the material required for the current task.

Related guides

Open Quickstart: Create Your First AI Partner and Verify a Task in the Docs navigation for the full creation flow. For conversation setup, open Connect and Manage Channels for an AI Partner.