issuing

Creates GitHub issues written as specs using the repository's own labels, types, and project fields.

1|Updated Sep 10, 2026
One-click install
npx skills add https://github.com/blauwtje/exo --skill issuing-blauwtje
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: issuing
Source: https://github.com/blauwtje/exo/tree/main/skills/issuing
Command: npx skills add https://github.com/blauwtje/exo --skill issuing-blauwtje

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve? Filing GitHub issues by hand produces story-style write-ups that repeat metadata, lack acceptance criteria, and ignore the labeling scheme the repository already uses, leaving later planning work without a spec to build against. ## Core Features & Use Cases - Spec-shaped issue bodies: Writes every issue in a Spec or Report structure with outcome, done-when criteria, and references, in the language of the repository's existing issues. - Repository vocabulary detection: Runs a script that reads labels, issue types, milestones, and project fields through the gh CLI, reusing what exists and creating default labels only when the repository defines none. - Deterministic sizing: Derives size, estimate, and priority from path and acceptance-criteria counts so two sessions produce the same values. - Use Case: Ask to split a feature request into a parent issue plus sub-issues; the skill grounds references in real code paths, creates the issues in dependency order with labels and project fields set, and reads each one back with its URL. ## Quick Start Ask the assistant to file a GitHub issue describing the outcome you want, for example to open an issue for adding rate limiting to the API endpoints.

Frequently Asked Questions about issuing

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

FAQPage Schema
How do I create GitHub issues from a plain-language request?▼

Describe the outcome the issue should cover and the skill turns it into a goal sentence, writes a Spec or Report body, and creates the issue with gh issue create. Labels, type, milestone, and project fields are set from the repository's own vocabulary.

How are GitHub issue labels and project fields chosen automatically?▼

A script reads the repository's labels, issue types, milestones, and project fields through the gh CLI and caches the result for a day. If the repository has its own vocabulary only those names are used; otherwise a default set of type, size, and priority labels is created.

Can this skill close or edit an existing GitHub issue?▼

No, it only creates issues and never closes, deletes, or edits existing ones. The single exception is adding a parent or blocker relation the user explicitly named; other changes are reported back for the user to run the gh command themselves.

What tools are required to file issues with this workflow?▼

It requires the GitHub CLI (gh), git, and Node.js on PATH, and the working directory must point at the target repository. The repo-fields script shells out to gh for all repository reads and writes.

Why does issue creation report unread fields?▼

Unread entries mark fields a gh call failed to read, such as projects when the token lacks read:project scope, rather than fields the repository lacks. Each unread entry is named in the report so a skipped read is never mistaken for an absent field.