> ## Documentation Index
> Fetch the complete documentation index at: https://docs.brightdata.com/llms.txt
> Use this file to discover all available pages before exploring further.
# Scraper Studio Auto Self-Healing
> Configure Auto Self-Healing in Bright Data Scraper Studio to fix a scraper automatically when its success rate drops or 1 of 6 supported errors occurs.
Use the **Auto Self-Healing** tab to let Bright Data Scraper Studio attempt to fix a scraper on its own when collection quality drops or a supported scraper error occurs.
Auto Self-Healing is configured per scraper. Once enabled, Bright Data Scraper Studio monitors that scraper's runs and starts a healing job whenever the trigger rules you set are met.
Auto Self-Healing runs without you asking. To fix a scraper on demand from a plain-language prompt instead, use the [Self-Healing tool](/datasets/scraper-studio/self-healing-tool) in the Scraper Studio IDE.
## How do I open Auto Self-Healing settings?
1. Open **My Scrapers** in Bright Data Scraper Studio.
2. Select the scraper you want to configure.
3. Open the **Auto Self-Healing** tab.
4. Configure the general settings, trigger rules and notifications.
5. Click **Save**.
Saved settings apply to future runs of that scraper. Runs that already completed are not reprocessed.
## Which general settings control Auto Self-Healing?
| Setting | Options | What it does |
| ---------------------------- | ------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Enable Auto self healing** | On, Off | Turns Auto Self-Healing on or off for this scraper. When On, Bright Data Scraper Studio can start healing jobs as soon as the trigger rules are met. |
| **Auto save code** | On, Off | Controls whether Bright Data Scraper Studio saves the code the healing process generates. When On, healed code is saved automatically once the healing job succeeds. When Off, the healing job still runs but the generated code is not saved, and the users you select under **Notifications** get an email so they can review the suggested fix before saving it. |
**Auto save code** applies AI-generated code to the scraper with no human review. Turn it on only when you are willing to let Bright Data Scraper Studio change scraper code unattended. Leave it Off to keep a review step before any fix reaches the scraper.
## Which trigger rules start a healing job?
Trigger rules define when Auto Self-Healing starts. All four rules apply to the single scraper you are configuring.
| Trigger rule | What it controls |
| --------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Success Rate Threshold (%)** | The minimum success rate allowed before Auto Self-Healing is triggered. Set it to `80` and Auto Self-Healing can start once the scraper's success rate falls below 80 percent. |
| **Trigger Frequency** | How often Bright Data Scraper Studio checks whether Auto Self-Healing should run. Supported values are **Per day** and **Per job**. |
| **Cooldown Period (hours)** | The minimum number of hours Bright Data Scraper Studio waits between Auto Self-Healing attempts for the same scraper. Use it to stop repeated healing attempts in a short window. |
| **Max Auto-healing Jobs per day** | The maximum number of Auto Self-Healing jobs that can run for this scraper in one day. Use it to cap how often automatic fixes are attempted. |
## Which error codes trigger Auto Self-Healing?
Auto Self-Healing starts only for these six scraper error codes:
* `crawl_error`
* `parse_error`
* `bad_cmd_arg`
* `dead_page`
* `bad_input`
* `bad_navigate`
A scraper that fails with any other error code does not start Auto Self-Healing, even when every trigger rule is met. For what each code means and how to resolve it by hand, see [Scraper Studio error codes](/datasets/scraper-studio/error-codes).
## Where do I see status and notifications?
The **Status** section of the **Auto Self-Healing** tab shows the current state for that scraper: whether Auto Self-Healing is enabled, whether a healing job is running now and whether the most recent attempt succeeded or needs review.
The **Notifications** section selects which users receive email about Auto Self-Healing activity. Selected users are notified when Auto Self-Healing is triggered, when it completes and when it needs attention. Add at least one user when **Auto save code** is Off, because the review email is how a suggested fix reaches a person.
## FAQ
### How is Auto Self-Healing different from the Self-Healing tool?
Auto Self-Healing runs on its own against trigger rules you configure per scraper. The [Self-Healing tool](/datasets/scraper-studio/self-healing-tool) is manual: you open a scraper in the Scraper Studio IDE, describe the fix in plain language and review the diff before accepting it. Auto Self-Healing needs no prompt.
### Does Auto Self-Healing change my scraper without asking?
Only when **Auto save code** is On. With **Auto save code** Off, Bright Data Scraper Studio still runs the healing job but does not save the generated code, and emails the users selected under **Notifications** to review the suggested fix first.
### Why did Auto Self-Healing not start after my scraper failed?
Check the failure's error code first. Auto Self-Healing starts only for the six error codes listed on this page, so a scraper that fails with any other code never triggers it. If the code is supported, check whether the scraper is still inside its **Cooldown Period** or has already hit **Max Auto-healing Jobs per day**.
### Do the settings affect runs that already finished?
No. Auto Self-Healing settings apply to future runs of the scraper. Runs that already completed are not changed or reprocessed.
## Related
Fix a scraper on demand from a plain-language prompt
What each scraper error code means and how to resolve it
Start a self-healing job from the AI Flow API
Run a production scraper and configure delivery