investigate-sentry-issue

name	investigate-sentry-issue
description	Investigate and triage a Sentry error issue Use when this capability is needed.
metadata	{"author":"desplega-ai"}

Investigate Sentry Issue

Investigate a Sentry issue to understand the error, gather context, and prepare for fixing or triaging.

Prerequisites

SENTRY_AUTH_TOKEN and SENTRY_ORG must be set. Verify with sentry-cli info.

Arguments

• sentry-issue-url-or-id: A Sentry issue URL or just the issue ID (e.g., 123456)

Workflow

1. Parse the Input

If given a URL (https://sentry.io/organizations/{org}/issues/{issue_id}/), extract the issue ID.

2. Get Issue Overview

sentry-cli issues list --id <issue-id>

3. Get Detailed Info via REST API

Use https://sentry.io/api/0/organizations/${SENTRY_ORG}/issues/<issue-id>/ for metadata (first/last seen, event count, user impact, status).

4. Get Full Stacktrace

Use the recommended event endpoint: .../issues/<issue-id>/events/recommended/

Key jq paths for the response:

• .entries[] | select(.type == "exception") — exception with stacktrace
• .entries[] | select(.type == "exception") | .data.values[].stacktrace.frames — stack frames
• .entries[] | select(.type == "breadcrumbs") — actions leading to the error
• .tags — environment, browser, OS info
• .context — custom context data

5. Analyze and Report

1. Identify the failing file and line number
2. Understand the error type and message
3. Review breadcrumbs for the sequence of events
4. Check tags for environment-specific issues

6. Take Action (if appropriate)

sentry-cli issues resolve <issue-id>    # Resolve
sentry-cli issues mute <issue-id>       # Mute
sentry-cli issues unresolve <issue-id>  # Unresolve

Search Query Syntax

Use these filters with sentry-cli issues list --query:

Filter	Example	Description
is:	is:unresolved	Issue status
lastSeen:	lastSeen:-2d	Seen within time range
message:	message:undefined	Match error message
issue.category:	issue.category:error	Error category

Tips

• Always get the recommended event first — it's curated for debugging
• Check breadcrumbs to understand user actions before the error
• If investigating from a Slack alert, the issue URL is in the alert message

---

Source: desplega-ai/agent-swarm — distributed by TomeVault.