You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Commit cd7cf7b
Browse filesBrowse the repository at this point in the historyBrowse files
metadesc: "The Analyzer Agent in Testsigma is your intelligent assistant that autonomously reviews test executions | This article discusses the Analyzer Agent in Testsigma"
4
+
metadesc: "The Analyzer Agent in Testsigma analyzes a failed test step and returns the error type, root cause, visual evidence, and suggestions | This article discusses the Analyzer Agent in Testsigma"
5
5
noindex: false
6
6
order: 4.815
7
7
page_id: "analyzer-agent-in-testsigma"
@@ -25,79 +25,79 @@ contextual_links:
25
25
26
26
---
27
27
28
-
The Analyzer Agent in Testsigma analyzes failed test steps and provides actionable insights. It identifies the error type, determines the root cause, provides visual evidence captured during execution, and suggests possible fixes. This article discusses how to analyze step failures and take action using the Analyzer Agent.
28
+
The Analyzer Agent analyzes a failed test step and returns the error type, the root cause, the visual evidence captured during execution, and a set of suggestions. Apply a suggestion to update the step, or report the failure to your bug tracking tool, without leaving the results page.
29
29
30
30
---
31
31
32
32
> <pid="prerequisites">Prerequisites</p>
33
-
>
33
+
>
34
34
> Before you begin, ensure that:
35
-
> 1. At least one test plan has been executed.
36
-
> 2. The test plan contains failed test cases.
37
-
35
+
> 1. At least one test plan or dry run has been executed.
36
+
> 2. The run contains a failed test step.
38
37
39
38
---
40
39
41
40
## **Steps to Analyze Step Failures**
42
41
43
42
1. From the left navigation bar, click **Run Results**.
2. Select the test plan that contains failed test steps.
44
+
2. Select the run that contains failed test steps.
47
45
48
46
3. Open the test case with failures and select the failed test step.
49
47
50
-
4. In the **Step Details** panel, click **Analyze with Agent**. This will open the **Analyzer Agent** overlay.
51
-
48
+
The step header shows the error message. Click **Read more** to see the full message. The **Analysis** tab shows the error code, such as **#NO_SUCH_ELEMENT**, along with the **Visual Evidence** section.
49
+
50
+
4. Click **Analyze with Agent** in the action bar at the bottom of the page.
5. Wait for the **Analyzer Agent** to complete the analysis.
53
+
5. Review the analysis, which displays:
54
+
<!-- Image Placeholder: Analysis -->
55
+
-**Error Type**: The category of failure identified during execution, such as ELEMENT_NOT_FOUND.
56
+
-**Root Cause**: An explanation of why the step failed, based on the error message and the captured evidence.
57
+
-**Visual Evidence**: The screenshot captured at the time of execution. Click **Expand Images** to view it at full size, or the download icon to save it.
58
+
-**Suggestions**: A numbered list of ways to resolve the failure.
54
59
55
-
6. Review the analysis in **Analyzer Agent** overlay, which displays:
-**Error Type**: Displays the category of failure identified during test execution.
58
-
-**Root Cause**: Describes the underlying reason for the failure based on execution analysis.
59
-
-**Visual Evidence**: Displays screenshots captured during recording and execution to support failure analysis.
60
-
-**Suggested Fixes**: Displays recommended actions to resolve the failed test step.
60
+
[[info | **NOTE**:]]
61
+
| To ask a follow-up question about the failure, enter it in **Debug with Atto** and click **Analyze**.
62
+
63
+
**Additional Information**: Responses are generated by Atto AI. Review the analysis before acting on it. Use the feedback icons or **Give Feedback** to tell us how useful it was.
61
64
62
65
---
63
66
64
67
## **Apply Fix Using Analyzer Agent**
65
68
66
-
1. From the **Analyzer Agent** overlay, select the fix you want to apply.
3. Wait for the **Analyzer Agent**to generate the required fix.
77
+
Atto generates the fix and returns an **Update Test Step**card explaining what it found, followed by the current version of the step and the proposed version.
73
78
74
-
4. Click **Add Step**, **Update Step**, or **Update Step Settings**, as suggested by the **Analyzer Agent**.
75
-
79
+
3. Compare the two versions of the step.
76
80
77
-
5. Verify that the changes are applied to the test step.
78
-
81
+
4. Click **Update Step** to apply the change.
79
82
80
-
6. Rerun the test plan to validate the fix.
83
+
5. Rerun the test to validate the fix.
81
84
82
85
---
83
86
84
87
## **Report a Bug Using Analyzer Agent**
85
88
86
-
1. From the **Analyzer Agent** overlay, click **Report Bug**.
- To raise a new issue, open the **Create New** tab, review the prefilled details, and click **Report Bug**.
96
+
- To attach the failure to an issue that already exists, open the **Link To Issue** tab, search for the issue, and click **Link To Ticket**.
99
97
100
98
[[info | **NOTE**:]]
101
-
| To link the failure to an existing issue, click **Link to Issue**, search for the issue in the selected bug tracking application, and then click **Link to Ticket**.
99
+
| The fields displayed vary based on the selected bug tracking tool.
102
100
103
-
---
101
+
**Additional Information**: If the tool returns no projects, check that the credentials configured for it have access to the projects you expect. For Jira, the panel reads "For given JIRA Credentials, the projects list is empty. Please make sure it has the right access to required projects on JIRA".
metadesc: "Review Auto-Heal Insights in Testsigma to see which elements were healed, how locators were updated, and decide if the changes should be made permanent."
4
+
metadesc: "Review how auto-healing behaved during a run in Testsigma | See which locator failed, what replaced it, and decide whether to keep the change."
5
5
noindex: false
6
6
order: 4.918
7
7
page_id: "auto-healing-insights"
@@ -10,32 +10,109 @@ contextual_links:
10
10
- type: section
11
11
name: "Contents"
12
12
- type: link
13
-
name: "Steps to View Auto-Heal Insights"
14
-
url: "#steps-to-view-auto-heal-insights"
13
+
name: "Prerequisites"
14
+
url: "#prerequisites"
15
+
- type: link
16
+
name: "Identify Healed Steps in a Run"
17
+
url: "#identify-healed-steps-in-a-run"
18
+
- type: link
19
+
name: "Review the Heal on the Analysis Tab"
20
+
url: "#review-the-heal-on-the-analysis-tab"
21
+
- type: link
22
+
name: "View Autoheal Details"
23
+
url: "#view-autoheal-details"
24
+
- type: link
25
+
name: "Read the Locator Trace"
26
+
url: "#read-the-locator-trace"
27
+
- type: link
28
+
name: "When a Heal Fails"
29
+
url: "#when-a-heal-fails"
30
+
- type: link
31
+
name: "Fix the Locator Yourself"
32
+
url: "#fix-the-locator-yourself"
33
+
---
34
+
35
+
---
36
+
37
+
Review how auto-healing behaved during a run to see which locator failed, what replaced it, and whether to keep the change. Insights open from the healed step on the results page.
38
+
39
+
---
40
+
41
+
> <pid="prerequisites">Prerequisites</p>
42
+
>
43
+
> Before you begin, ensure that:
44
+
> 1. Auto-healing is enabled for the project.
45
+
> 2. A run has been completed in which at least one step was healed.
| A healed step reports **Passed**, because the step completed. The run itself carries a **Healed** status in the run list, which distinguishes it from a run that passed without intervention.
58
+
15
59
---
16
60
61
+
## **Review the Heal on the Analysis Tab**
62
+
63
+
The **Analysis** tab shows a **Healed this step** card explaining what happened: "Existing locator failed during execution, and we've auto-healed it using AI. The element locator was updated with a new XPath to ensure your test continues to run smoothly."
- Click **Approve as Primary** to make the healed locator the locator the test uses from now on.
68
+
- Click **Ignore** to leave the original locator in place. The heal still applied to this run only.
69
+
70
+
[[info | **NOTE**:]]
71
+
| In the **Visual Evidence** section, the current run panel carries a **Healed** badge, and the **Element** field has an edit icon for correcting the locator directly.
72
+
73
+
---
74
+
75
+
## **View Autoheal Details**
76
+
77
+
Open the **Autoheal Details** panel to see exactly which locator changed.
- The prompt "Do you want to update the autohealed element in all the linked test cases?" with **Update** and **Ignore** buttons.
83
+
- The healing method and how long it took, such as "Autohealed using CSS Selector" and "32s 912ms".
84
+
- The locator that failed, struck through, followed by the locator that replaced it.
85
+
86
+
Click **Update** to apply the healed locator across every test case linked to the element, or **Ignore** to leave those test cases unchanged.
87
+
88
+
**Additional Information**: Use the feedback icons under "Is this autoheal helpful?" to tell us whether the heal was correct. The same feedback icons appear on the auto-healed banner above the step list.
89
+
17
90
---
18
91
19
-
Once you’ve executed your tests using Auto-Heal, you can review detailed insights to understand how the feature worked during execution. The Auto-Heal Insights section helps you identify which elements were healed, how their locators were updated, and whether any of those updates should be made permanent.
92
+
## **Read the Locator Trace**
93
+
94
+
Every auto-heal event records a locator trace: the full sequence the engine worked through to arrive at a heal, and whether that heal succeeded or failed.
20
95
21
-
This article discusses how to view and manage Auto-Heal insights after test execution.
96
+
Use the trace to understand why a particular locator was chosen, or why no replacement could be found.
97
+
98
+
**Additional Information**: The locator trace is provided by the Auto Heal V2 architecture.
22
99
23
100
---
24
101
25
-
## **Steps to View Auto-Heal Insights**
102
+
## **When a Heal Fails**
103
+
104
+
A heal attempt that fails gets the same root cause treatment as a failed step. Open the **Root cause** block on the **Analysis** tab and click **Explain this failure** to see why the engine could not resolve the element.
26
105
27
-
1. Click the test step where the element was auto-healed to view insights.
106
+
[[info | **NOTE**:]]
107
+
| This requires Analyzer V2. Without it, a failed heal reports that it did not succeed, but no explanation is generated.
28
108
29
-
2. Go to the **Step Result** and click **Auto-Heal Details**. This opens the **Auto-Heal Details** overlay.
-**Update element** corrects the locator directly. In the **Visual Evidence**section, the**Element**field on the current run side carries an edit icon for this.
116
+
-**Relearn step** re-captures the step so a fresh locator is recorded.
0 commit comments