skills.rest
kortix-ai
kortix-aiOFFICIAL
Explore all skills by this author →
Website𝕏@kortixGitHub

error-backlog-triage

Groups and dedupes Sentry errors into drafted GitHub issues with Slack summaries.

20.2k|3.4k|Updated Oct 5, 2024
Search Tags:#deduplication#Sentry#backlog grooming#error triage#gitHub issues#slack alerts#Monitoring
One-click install
npx skills add https://github.com/kortix-ai/suna --skill error-backlog-triage
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: error-backlog-triage
Source: https://github.com/kortix-ai/suna/tree/main/packages/starter/templates/marketplace/runtime/skills/error-backlog-triage
Command: npx skills add https://github.com/kortix-ai/suna --skill error-backlog-triage

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Engineering teams accumulate noisy Sentry backlogs where new and spiking errors go unnoticed or get reported as duplicate GitHub issues. This Skill runs a recurring, read-only grooming sweep that turns the raw error backlog into a short, ranked, deduplicated set of drafted issues without paging anyone or mutating Sentry state.

Core Features & Use Cases

  • Backlog grouping and ranking: Pulls new and spiking errors from Sentry, groups them by fingerprint, and ranks them by an impact score combining event count and affected users.
  • GitHub dedupe and drafting: Checks each candidate against existing issues in the triage repo by Sentry issue ID, then drafts labeled GitHub issues with stack traces and impact data for the top offenders.
  • Sweep summary alerts: Posts one Slack summary per sweep listing drafted issues and already-tracked candidates, staying silent when nothing is new.
  • Use Case: A platform team schedules an hourly sweep so that every morning the backlog is already groomed — the three highest-impact errors each have a drafted GitHub issue with full stack traces, and duplicates link back to existing tickets.

Quick Start

Ask the agent to groom the Sentry backlog for your project and draft GitHub issues for the top new or spiking errors, then post a summary to your alert channel.

Frequently Asked Questions about error-backlog-triage

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automatically create GitHub issues from Sentry errors?▼

Run a sweep that pulls new and spiking Sentry issues, groups them by fingerprint, and drafts GitHub issues via the gh CLI with the stack trace, impact counts, and Sentry permalink embedded in the body. Each issue is labeled bug and sentry-triage with no assignee.

How to avoid duplicate GitHub issues for the same Sentry error?▼

Search the triage repo with gh issue list for the Sentry issue short ID or permalink before drafting. Issues drafted by this workflow embed the Sentry ID in the body, so matches are marked already tracked and skipped instead of re-drafted.

Does this workflow resolve or mute Sentry issues automatically?▼

No, the sweep is strictly read-only against Sentry. It never resolves, ignores, mutes, merges, or deletes issues, and drafted GitHub issues carry no assignee — a human owns triage and prioritization from there.

How are spiking errors detected versus new errors?▼

An error counts as spiking when its event frequency over the last hour reaches a configurable multiplier of its trailing baseline, such as the 14-day hourly average. Spiking and new errors are ranked together on the same impact score of event count times affected users.

What happens when more errors qualify than the per-run limit?▼

Only the top candidates up to the configured max_issues_per_run are drafted in a single sweep. Remaining candidates are not queued or remembered; the next sweep re-evaluates them against the current backlog state.

Related Skills

Explore More Software Engineering Skills→
dr-codedr-code

sentry-automation

Automate Sentry issue, alert, and release management across organizations and projects.

Community
Intermediate
kaminocorpkaminocorp
228

github-issues

Manage GitHub issue lifecycles across repositories using gh CLI or curl.

Official
Intermediate
Srujan0798Srujan0798

incident-response

Coordinates end-to-end incident response workflows including triage, communication, and postmortems.

Community
Advanced
GainInsightDevGainInsightDev

af-integrate-sentry

Integrate Sentry error tracking into Next.js, Node, React, and Python projects.

Official
Advanced