triage

Processes work items into GitHub issues via automated research and routing.

Updated Apr 7, 2026

Search Tags:#automation #research #routing #classification #triage #batch-processing #github-issues

Ready to run

npx skills add https://github.com/bzellman/earp-kit --skill triage-bzellman

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: triage
Source: https://github.com/bzellman/earp-kit/tree/main/skills/system/triage
Command: npx skills add https://github.com/bzellman/earp-kit --skill triage-bzellman

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Batch-process loosely described work items into properly-routed GitHub issues by autonomously researching each item against the codebase, classifying by size/confidence/risk, and routing to the appropriate path.

Core Features & Use Cases

Research items in parallel against the codebase to determine affected files, scope, complexity, and unknowns.
Classify each item by size (small/medium/large), confidence (high/medium/low), type (bug/enhancement), and risk (low/medium/high).
Route items to [Research], [Ready], or [REVIEW], and auto-create GitHub issues with appropriate labels and content.

Quick Start

Paste a batch of loosely described items into the triage tool and have it generate properly labeled GitHub issues through automated research and routing.

Frequently Asked Questions about triage

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

FAQPage Schema

How do I automatically batch process loosely described ideas into GitHub issues?▼

To batch process loosely described ideas into GitHub issues, you can paste a list of raw work items into a triage tool. The system researches each item against your codebase, classifies it by size, confidence, and risk, and generates properly routed issues with appropriate labels.

How does autonomous triage classify and route work items in software engineering?▼

Autonomous triage classifies items by size (small/medium/large), confidence (high/medium/low), type (bug/enhancement), and risk (low/medium/high). It then routes items to categories like Research, Ready, or REVIEW based on the automated codebase research findings.

Can I use automated triage to research multiple bugs and features in a single message?▼

Yes, automated triage is designed for teams that receive multiple bugs, features, and ideas in a single message. It parses the batch input and performs parallel exploration of your codebase to determine affected files, scope, and complexity for each item.

What is the best way to turn a raw list of features and bugs into structured GitHub issues?▼

The best way to turn raw lists into structured GitHub issues is using an automated triage process. It parses input, autonomously researches the codebase to identify unknowns, and auto-creates issues with structured content, scope labels, and routing paths.

Does automated issue routing require manual classification of risk and scope?▼

No, automated issue routing does not require manual classification. The system autonomously assigns labels for scope, type, size, confidence, and risk by researching the codebase, then routes the generated issues to the appropriate development path.

When should I not use automated batch triage for GitHub issue generation?▼

You should avoid automated batch triage for single, highly detailed work items that already have defined scopes and routing paths. This system is optimized for batch-processing multiple loosely described items that require parallel codebase research and classification.