datawebotRun datawebot on a schedule
Have your agent set up scheduled datawebot reads with cron, GitHub Actions or its own scheduler, and keep the results in your own records.
Ask your agent to set it up
The quickest route is to tell your agent what you want checked and how often. Paste this into a coding agent such as Claude Code or Cursor, and replace the part in brackets. Product links work best, because the agent then has nothing to search for. If you are new to datawebot, the first line signs you in with an email approval.
set up https://mcp.datawebot.com/setup.md
Then write a script that reads [these product links, or this store] every day at 9:00 and saves each result with the time it was read. Have it read its key from DATAWEBOT_API_KEY, and tell me how to create one on the dashboard. Tell me the credits it will use each month, then schedule it with cron or GitHub Actions.Choose where it runs
| Route | Runs on | Use it when |
|---|---|---|
| cron | A computer or server that stays on | You already have a machine running |
| GitHub Actions | GitHub, on a timer | You have no machine to leave on; start times can slip at busy hours |
| Your agent's scheduler | Your agent, if it can run tasks on a timer | You want the agent to read the results and report back |
Give the schedule a key
A scheduled run has nobody there to approve a sign-in, so give it its own key. The key your agent saves when you sign in lasts 30 days and has a daily limit of 100 credits, which suits everyday use but not a schedule. Open API keys on your dashboard, create a key that expires in 365 days, and store it as DATAWEBOT_API_KEY in your environment or as a GitHub secret. Replace it before it expires.
Know the cost
A scheduled read costs the same as any other: 1 credit for a product read and 5 for a search, and failed requests are free. Checking 100 products once a day uses about 3,000 credits in 30 days. Pass max_credits to cap what any one request can cost.
Keep your own records
Every answer says when it was read, so saving each run gives you a dated history in your own database or spreadsheet. Compare the latest run with the one before it to see what changed.
- For a complete script with SQLite, cron and GitHub Actions, read the price tracker tutorial at /blog/price-tracker-python.
- For the use case, see /solutions/price-monitoring-api.
Need help? Contact support.