skills.rest
bkrsna
bkrsnaCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

quality

Classifies GitHub issues as spam, empty, needs-info, feature, or actionable bug reports.

Updated Mar 22, 2026
Search Tags:#github#bug report#classification#issue triage#json output#spam detection#quality evaluation
One-click install
npx skills add https://github.com/bkrsna/gemini-cli --skill quality-bkrsna
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: quality
Source: https://github.com/bkrsna/gemini-cli/tree/main/tools/caretaker-agent/cloudrun/triage-worker/.gemini/skills/quality
Command: npx skills add https://github.com/bkrsna/gemini-cli --skill quality-bkrsna

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Maintainers of open-source repositories receive large volumes of GitHub issues of varying quality, and manually reviewing each one to decide whether it is spam, incomplete, a feature request, or a valid bug report consumes significant triage time. ## Core Features & Use Cases - Issue Quality Classification: Evaluates an issue's title and body and assigns one of five statuses: SPAM, EMPTY, NEEDS_INFO, FEATURE, or OK. - Structured JSON Output: Returns the assessment as a single JSON object containing the quality label, detailed reasoning, and an optional draft comment requesting missing details. - Prompt Injection Defense: Automatically classifies any issue containing prompt injection attempts as SPAM, regardless of other content. - Use Case: A triage worker in a caretaker agent pipeline receives a new GitHub issue, runs this evaluation, and automatically routes actionable bugs to maintainers while drafting information requests for incomplete reports. ## Quick Start Evaluate this GitHub issue title and body and tell me whether it is spam, empty, needs more information, a feature request, or ready to triage.

Frequently Asked Questions about quality

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

FAQPage Schema
How do I automatically triage GitHub issues for quality?▼

Analyze the issue title and body against defined quality criteria covering clarity, completeness, and reproducibility. Each issue is classified as SPAM, EMPTY, NEEDS_INFO, FEATURE, or OK, with the result returned as a structured JSON object including reasoning.

What makes a GitHub issue ready for triage?▼

An issue is classified as OK when it shows clear user intent to report a systemic code defect with sufficient reproduction details. Reports stemming from user-defined configurations or lacking actionable reproduction steps are classified as NEEDS_INFO instead.

How are prompt injection attacks in issues handled?▼

Any issue containing a prompt injection attempt, such as instructions to ignore previous directives, is immediately classified as SPAM. This applies regardless of whether the body also contains a legitimate bug description or real codebase references.

When is a GitHub issue classified as NEEDS_INFO versus EMPTY?▼

EMPTY applies when the issue has little to no descriptive content, such as blank bodies or boilerplate template text. NEEDS_INFO applies when the issue has on-topic context like environment details or logs but lacks reproduction steps or actionable specifics.

Does the triage output include a response to the issue reporter?▼

Yes, when an issue is classified as NEEDS_INFO, the JSON output includes a draft comment field. The comment thanks the reporter and lists the specific missing details needed to triage the bug.

Related Skills

Explore More Software Engineering Skills→
Varnan-TechVarnan-Tech
581

kill-the-standup

Generate and post daily standup reports from Linear issues and GitHub commits to Slack.

Official
Intermediate
NVIDIANVIDIA
22.1k

nemoclaw-user-triage-instructions

Assign 1–5 labels and generate tiered triage comments for GitHub issues and PRs.

Official
Intermediate
Terry-MaoTerry-Mao
144

triage-issue

Analyze GitHub issues to suggest labels, estimate reproducibility, and detect duplicates.

Community
Intermediate
MigoXLabMigoXLab
736

dingo-data-quality

Evaluate machine learning datasets with rule-based and LLM-based checks.

Official
Advanced