Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.sequentum.com/llms.txt

Use this file to discover all available pages before exploring further.

This walkthrough uses Control Center. For the API version, see Quickstart.
1

Open a new session

From your Space, click New agent → Build with AI.
2

Describe what you want

Type a prompt like:
Pull product name, price, image URL, and SKU from https://example-store.com/products/. Skip out-of-stock items.
The more concrete the prompt, the fewer follow-ups you’ll need.
3

Watch the plan

Agent Builder reports each phase as it works. You’ll see a live preview of the pages it visits and the records it pulls.
4

Open the agent

When the build completes, click Open Agent. You land on the agent details screen — the same screen any agent in your Space uses.
5

Run it

From agent details, run the agent manually, attach a Schedule, or hit POST /agent/{agentId}/start from your code.

​
Tips for better prompts

At launch, Agent Builder works best as a single-shot experience: describe what you want comprehensively in your first prompt and the agent is built and ready to run. Iterative refinement (follow-ups inside the same session) is on the roadmap. Until then, treat the first prompt as your one shot — be specific.
“Extract X, Y, Z from https://...” outperforms “scrape this site for me.”
Say “follow all paginated pages” or “first 50 pages only” — Agent Builder will scope the run.
If the site needs a login, mention it. Agent Builder will collect the credentials before it starts navigating.
“One row per product with columns X, Y, Z” beats “give me the data.” Naming the schema in plain language guides the extractor.
“Skip out-of-stock items.” “Ignore the related-products section.” Negative scoping prevents noise.