Step 1
Record the merchant's stated goals, the actions permitted to run unattended, the prohibited actions, the caps on volume, spend and affected records per run and per period, the review cadence, and the reviewer named for it.
The merchant is willing to let an agent act on its own, but only within permissions and caps that are written down. In this case, the immediate task is to design an autonomous operations loop with clear action limits. The workflow should produce an agent task plan and action log showing each autonomous action, its justification and its review status.
First replace every [bracketed] placeholder with real store scope and constraints, then add any actual records or data you have, and paste it into your preferred AI assistant or Spark agent. Spark will follow this page's workflow to produce structured output and verification steps.
Act as a Shopify operations specialist.
Help me design an autonomous operations loop with clear action limits.
My store context is [describe the business, products and target audience].
My scope is [specify the affected products, pages, customers or reporting period]. Use these inputs: [the merchant's stated goals, the actions permitted to run unattended, the prohibited actions, per-run and per-period caps on volume, spend and affected records, the review cadence and its named reviewer]. Bound the loop so every self-directed step stays inside those permissions and caps, pauses when corroborating evidence is thin or two permitted actions conflict, and reverts any action whose verification reading does not match the intended outcome.
Return an agent task plan and action log showing each autonomous action, its justification and review status for the specified task.
Show the relevant evidence or before-and-after examples and explain how I can verify the result.
If essential data is missing, ask for it and mark unsupported conclusions as unknown.
Do not invent metrics, product claims or customer facts.
For any publication, message, account or store change, use only connected tools within my explicitly authorized scope and report success only after verification; otherwise provide the draft or proposed steps.Split the work into input, judgment, and deliverable phases. Verify each step's output before moving to the next.
Record the merchant's stated goals, the actions permitted to run unattended, the prohibited actions, the caps on volume, spend and affected records per run and per period, the review cadence, and the reviewer named for it.
Bound the loop so every self-directed step stays inside those permissions and caps, pauses when corroborating evidence is thin or two permitted actions conflict, and reverts any action whose verification reading does not match the intended outcome.
Deliver the task plan and action log showing per autonomous action its trigger, justification, the cap it consumed, its verification result and review status, and count every pause and revert so the merchant can see where the loop stopped itself.
Cover decision rules, frequent misunderstandings, and required inputs so the team aligns on prerequisites before execution.
An enumerated prohibited-action list, caps on volume, spend and records touched per run and per period, a pause rule for thin evidence, and a named reviewer who reads the log; a cap with no owner behind it is simply unenforced.
No. Routine approvals shift onto predefined rules that the action log can evidence, while exceptional work exceeding a cap or touching a prohibited field still stops and waits for a named person.
Whenever evidence is thin, two permitted actions conflict or a cap is reached, and also after any action whose verification reading fails, since continuing would compound a change that was never confirmed.
Start with scenarios under the same topic, then expand into cross-topic matches suggested by similarity.
Ciwi Spark builds Shopify-native AI tools and agents around the workflows merchants actually run: research, ads, translation, and customer support. Stop rewriting the same prompts by hand.