> ## Documentation Index
> Fetch the complete documentation index at: https://mintlify.com/getsentry/cli/llms.txt
> Use this file to discover all available pages before exploring further.
# AI-Powered Debugging with Seer
> Use Seer AI to analyze root causes and generate solution plans for Sentry issues
Sentry CLI integrates with Seer AI to provide automated root cause analysis and solution planning for issues. The AI analyzes your error data, stack traces, and code context to identify what caused the issue and suggest how to fix it.
## Prerequisites
* GitHub integration configured for your organization
* Code mappings set up for your project
* Seer AI enabled in your organization settings
Seer AI is available on Business and Enterprise plans. Visit your organization's **Settings > Seer** to enable it.
## Commands
### Explain: Root Cause Analysis
The `issue explain` command analyzes an issue and identifies the root cause:
```bash theme={null}
sentry issue explain 123456789
```
The analysis provides:
* Identified root causes
* Reproduction steps
* Relevant code locations
````bash theme={null}
$ sentry issue explain 123456789
⠋ Analyzing issue...
## Root Cause Analysis Complete
### Cause #0: Null pointer exception in user authentication
**Repository:** my-org/my-app
**Reproduction steps:**
**Authentication flow when email is missing**
The code attempts to access the email property without checking
if the user object exists:
```python
def authenticate_user(user):
return user.email.lower() # user could be None
````
To create a plan, run: sentry issue plan 123456789
````
```bash
$ sentry issue explain 123456789 --json
````
````json theme={null}
[
{
"description": "Null pointer exception in user authentication",
"relevant_repos": ["my-org/my-app"],
"root_cause_reproduction": [
{
"title": "Authentication flow when email is missing",
"code_snippet_and_analysis": "The code attempts to access the email property without checking if the user object exists:\n\n```python\ndef authenticate_user(user):\n return user.email.lower() # user could be None\n```"
}
]
}
]
````
#### Issue ID Formats
You can reference issues in multiple ways:
```bash theme={null}
# Numeric ID
sentry issue explain 123456789
# Organization/short ID
sentry issue explain sentry/EXTENSION-7
# Project + suffix
sentry issue explain cli-G
# Short ID (searches across orgs)
sentry issue explain CLI-G
```
#### Force Fresh Analysis
Use `--force` to trigger a new analysis even if one exists:
```bash theme={null}
sentry issue explain 123456789 --force
```
### Plan: Solution Generation
The `issue plan` command generates a solution plan with implementation steps:
```bash theme={null}
sentry issue plan 123456789 --cause 0
```
The plan command automatically runs root cause analysis if needed, so you don't need to run `explain` first.
```bash theme={null}
$ sentry issue plan 123456789 --cause 0
Creating plan for cause #0...
"Null pointer exception in user authentication"
⠋ Generating solution...
## Solution
**Summary:** Add null check before accessing user.email property
### Steps to implement
1. **Add user validation in authenticate_user**
Update the authenticate_user function to check if user is None
before accessing the email property. Return a default value or
raise an appropriate exception.
2. **Add unit tests for null user case**
Create test cases to verify the function handles None users
correctly without throwing unhandled exceptions.
3. **Update caller code to handle exceptions**
Ensure all callers of authenticate_user are prepared to handle
the authentication failure cases appropriately.
```
```bash theme={null}
$ sentry issue plan 123456789 --cause 0 --json
```
```json theme={null}
{
"run_id": "autofix-abc123",
"status": "COMPLETED",
"solution": {
"one_line_summary": "Add null check before accessing user.email property",
"steps": [
{
"title": "Add user validation in authenticate_user",
"description": "Update the authenticate_user function to check if user is None before accessing the email property. Return a default value or raise an appropriate exception."
},
{
"title": "Add unit tests for null user case",
"description": "Create test cases to verify the function handles None users correctly without throwing unhandled exceptions."
},
{
"title": "Update caller code to handle exceptions",
"description": "Ensure all callers of authenticate_user are prepared to handle the authentication failure cases appropriately."
}
]
}
}
```
#### Selecting a Root Cause
If multiple root causes are identified, you must specify which one to plan for:
```bash theme={null}
# If you try without --cause when multiple exist:
$ sentry issue plan 123456789
Error: Multiple root causes found. Please specify one with --cause :
0: Null pointer exception in user authentication...
1: Missing rate limiting on authentication endpoint...
Example: sentry issue plan 123456789 --cause 0
```
Then specify the cause ID:
```bash theme={null}
sentry issue plan 123456789 --cause 0
```
#### Regenerate a Plan
Use `--force` to create a new plan even if one exists:
```bash theme={null}
sentry issue plan 123456789 --cause 0 --force
```
## Scripting with JSON Output
Both commands support `--json` for automation:
```bash Extract root causes theme={null}
#!/bin/bash
ISSUE_ID="123456789"
ROOT_CAUSES=$(sentry issue explain "$ISSUE_ID" --json)
COUNT=$(echo "$ROOT_CAUSES" | jq 'length')
echo "Found $COUNT root causes"
```
```python Generate plans for all causes theme={null}
import json
import subprocess
issue_id = "123456789"
# Get root causes
result = subprocess.run(
["sentry", "issue", "explain", issue_id, "--json"],
capture_output=True,
text=True
)
causes = json.loads(result.stdout)
# Generate plan for each cause
for i, cause in enumerate(causes):
print(f"Planning for cause {i}: {cause['description']}")
plan_result = subprocess.run(
["sentry", "issue", "plan", issue_id, "--cause", str(i), "--json"],
capture_output=True,
text=True
)
plan = json.loads(plan_result.stdout)
print(f"Solution: {plan['solution']['one_line_summary']}")
```
```javascript Node.js integration theme={null}
const { execSync } = require('child_process');
const issueId = '123456789';
// Get explanation
const explainOutput = execSync(
`sentry issue explain ${issueId} --json`,
{ encoding: 'utf-8' }
);
const causes = JSON.parse(explainOutput);
// Get plan for first cause
if (causes.length > 0) {
const planOutput = execSync(
`sentry issue plan ${issueId} --cause 0 --json`,
{ encoding: 'utf-8' }
);
const plan = JSON.parse(planOutput);
console.log('Solution:', plan.solution.one_line_summary);
}
```
## Troubleshooting
### Seer Not Enabled
If you see an error about Seer not being enabled:
```bash theme={null}
Error: Seer is not enabled for this organization.
Enable it at: https://sentry.io/settings/your-org/seer/
```
Visit your organization settings to enable Seer AI.
### No Budget Remaining
```bash theme={null}
Error: Your organization has exceeded its Seer budget for this billing period.
```
Contact your billing admin or wait for the next billing cycle.
### GitHub Integration Required
```bash theme={null}
Error: GitHub integration must be configured for your organization.
```
Seer requires GitHub integration and code mappings to analyze your code:
1. Install the GitHub integration in **Settings > Integrations**
2. Configure code mappings in your project settings
3. Try the command again
### Analysis Takes Too Long
The first analysis may take several minutes. Progress is shown with a spinner:
```bash theme={null}
⠋ Analyzing issue...
```
If it times out, try again - subsequent analyses use cached results and complete faster.