kurko/tool-sentry

Sentry Issue Lookup

Fetches a Sentry issue and its most representative event, then prints a structured debugging summary: error details, stacktrace (in-app frames first), HTTP request context, breadcrumbs, user info, and tags.

Prerequisites

The SENTRY_AUTH_TOKEN environment variable must be set.

If the token is missing, tell the user:

SENTRY_AUTH_TOKEN is not set. Create a token at https://sentry.io/settings/account/api/auth-tokens/ with event:read scope, then run:

export SENTRY_AUTH_TOKEN='your-token-here'

Do NOT attempt to run the script without the token.

Usage

# Fetch issue details (the full URL with query params is fine — only org and issue ID are used)
~/.claude/skills/tool-sentry/sentry-readonly-cli "https://myorg.sentry.io/issues/7044654032/"

# Show help with full output format documentation
~/.claude/skills/tool-sentry/sentry-readonly-cli help

Output Sections

The script outputs these sections (omitted if no data):

1. Header — short ID, title, status, level, culprit, event/user counts, timestamps, assignee
2. Error — error message from issue metadata
3. Request — HTTP method, URL, query string, body (if the error was from an HTTP request)
4. User — affected user's ID, email, username, IP
5. Contexts — runtime (e.g., Ruby 3.4.8), OS, browser
6. Stacktrace — in-app frames with source context first, then vendor/framework frames
7. Breadcrumbs — last 15 timestamped actions before the error (HTTP calls, DB queries, etc.)
8. Link — permalink to Sentry web UI
9. Tags — all event tags (environment, server, release, transaction, etc.)

Tips for AI Consumers

• The stacktrace separates in-app frames (your code) from vendor frames. Focus on in-app frames first.
• Breadcrumbs show the sequence of events leading to the crash — useful for understanding what the user/system was doing.
• Request context helps reproduce HTTP-triggered errors.
• Tags include the environment, release SHA, server name, and transaction — useful for scoping the issue.