issue

Create, update, find, and triage GitHub issues with templates and evidence-backed investigation.

4|1|Updated Jul 31, 2025
One-click install
npx skills add https://github.com/alvis/.agents --skill issue-alvis
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: issue
Source: https://github.com/alvis/.agents/tree/main/plugins/coding/skills/issue
Command: npx skills add https://github.com/alvis/.agents --skill issue-alvis

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Managing GitHub issues manually leads to duplicate reports, inconsistent formatting, unverified triage conclusions, and lost context when concurrent edits occur. This Skill standardizes issue creation, updating, lookup, and triage with repository templates, bounded search budgets, and read-back verification. ## Core Features & Use Cases - Issue Lifecycle Management: Create, update, look up, and triage GitHub issues using repository templates, existing labels, milestones, issue types, and project fields discovered via the GitHub API. - Duplicate Detection: Run bounded lexical searches across open and closed issues, rank candidates by behavioral match, and close proven duplicates with native duplicate-of relationships. - Evidence-Backed Triage: Delegate read-only code inspection, publish analysis comments with permalinks pinned to verified commits, and classify issues as Bug, Feature, or Task. - Use Case: A maintainer asks to triage the three highest-priority open issues; the Skill retrieves the queue, checks duplicates, inspects code at a verified revision, and posts verified classification comments. ## Quick Start Triage the top three open issues in this repository and post evidence-backed responses for each.

Frequently Asked Questions about issue

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

FAQPage Schema
How do I triage GitHub issues automatically?

The triage action retrieves open issues with priority-ordered queries, excludes already-investigated or waiting issues, runs duplicate lookup, classifies each as Bug, Feature, or Task, and posts a verified response comment. It processes one issue at a time within shared search budgets.

How do I find duplicate GitHub issues before creating one?

The lookup action searches open and closed issues with multiple focused queries via the GitHub search API, deduplicates candidates, and ranks them by behavioral match. Creation runs this lookup first and returns an established duplicate's URL instead of posting a new issue.

What tools are required to manage GitHub issues from the command line?

This workflow requires an authenticated GitHub CLI (gh), jq for JSON construction, and Git for inspecting code at specific revisions. It checks gh auth status for the target host and reports missing prerequisites without installing tools or broadening scopes.

Can it apply labels, milestones, and project fields to GitHub issues?

Yes, it discovers existing labels, milestones, issue types, and Projects V2 fields through the API and applies only values supported by repository conventions. Each metadata field is written separately and read back so partial failures are reported explicitly.

What happens if a GitHub issue write times out or fails?

On a timeout or unknown write result, the Skill rereads the target before retrying and reuses an existing equivalent comment or issue instead of duplicating it. At most two targeted retries are attempted, and unresolved outcomes are reported rather than claimed as success.

Does issue triage modify source code or run reproduction scripts?

No, triage treats issue content as untrusted data and never executes reproduction code or modifies source. Code inspection is delegated to a read-only analysis specialist, and fixes belong to separate coding skills.