sentry issue explain
Get a root cause analysis for a Sentry issue using Seer AI. This command analyzes the issue and provides identified root causes, reproduction steps, and relevant code locations.
Usage
sentry issue explain <issue-id>
explain command accepts the same issue identifier formats as view:
# Explicit org + ID
sentry issue explain sentry/EXTENSION-7
# Project + suffix
sentry issue explain cli-G
# Short ID (searches across orgs)
sentry issue explain CLI-G
# Suffix only (requires DSN context)
sentry issue explain G
# Numeric ID
sentry issue explain 123456789
Parameters
Issue identifier in any supported format.
Flags
Force a fresh analysis even if one already exists. Useful when issue context has changed.Example:sentry issue explain 123456789 --force
Output results as JSON array of root causes.Example:sentry issue explain EXTENSION-7 --json
Examples
Basic Usage
# Analyze issue by numeric ID
sentry issue explain 123456789
# Analyze by short ID
sentry issue explain EXTENSION-7
# With explicit org
sentry issue explain sentry/CLI-G
Force Re-analysis
# Trigger fresh analysis
sentry issue explain 123456789 --force
JSON Output
# Export to JSON
sentry issue explain EXTENSION-7 --json > root-causes.json
# Process with jq
sentry issue explain CLI-G --json | jq '.[0].description'
Sample Output
Analyzing issue... (this may take a few minutes)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Root Cause Analysis
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Root Cause #0:
The error occurs because the `processData` function attempts to call `.map()`
on a variable that is undefined. This happens when the API response does not
include the expected `users` array in the payload.
Reproduction Steps:
1. Navigate to /dashboard in the application
2. Ensure the external API endpoint returns a response without the 'users' field
3. The application attempts to render the user list
4. TypeError is thrown when calling users.map()
Relevant Code:
File: src/components/UserList.js
Line: 42
Code: const userElements = users.map(user => renderUser(user));
Tip: To create a plan, run: sentry issue plan 123456789
[
{
"id": 0,
"title": "Undefined variable access in processData",
"description": "The error occurs because the `processData` function attempts to call `.map()` on a variable that is undefined. This happens when the API response does not include the expected `users` array in the payload.",
"reproduction": [
"Navigate to /dashboard in the application",
"Ensure the external API endpoint returns a response without the 'users' field",
"The application attempts to render the user list",
"TypeError is thrown when calling users.map()"
],
"code_context": {
"file": "src/components/UserList.js",
"line": 42,
"snippet": "const userElements = users.map(user => renderUser(user));"
}
}
]
How It Works
- Trigger Analysis: The command sends the issue to Seer AI for analysis
- Poll Status: The CLI polls the analysis status every few seconds
- Extract Root Causes: Once complete, root causes are extracted from the analysis steps
- Display Results: Formatted output shows the identified causes with context
The analysis may take several minutes for complex issues. Progress is shown via a spinner (hidden in --json mode).
Requirements
Seer AI must be enabled for your organization. If Seer is not available, you’ll see an error with a link to enable it in your organization settings.
Prerequisites
- Seer AI enabled in organization settings
- Available Seer AI budget/quota
- Organization has AI features enabled
Common Errors
Seer not enabled:
Seer AI is not enabled for this organization.
Enable it at: https://sentry.io/settings/your-org/processing/ai-tools/
Seer AI analysis quota exceeded. Check usage at:
https://sentry.io/settings/your-org/processing/ai-tools/
AI features are disabled for this organization.
Notes
- The first analysis of an issue takes longer as Seer examines code context, stack traces, and event patterns.
- Subsequent runs reuse cached analysis unless
--force is specified.
- If multiple root causes are found, each is numbered (0, 1, 2, …) for reference in
sentry issue plan --cause N.
- Organization context is required to access Seer AI. The command automatically resolves org from the issue ID.
See Also
