.opencode/skill/gh-issue-triage/SKILL.md
GitHub issue triage workflow with contributor profile extraction. Analyze → clarify → file cells → tag → implement → credit. Captures Twitter handles for changeset acknowledgments.
npx skillsauth add joelhooks/swarm-tools gh-issue-triageInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
name: gh-issue-triage description: GitHub issue triage workflow with contributor profile extraction. Analyze → clarify → file cells → tag → implement → credit. Captures Twitter handles for changeset acknowledgments. tags:
Issues are conversations, not tickets. Treat contributors with respect - they took time to file the issue. Extract their profile info so changesets can properly credit them when fixes ship.
┌─────────────────────────────────────────────┐
│ ANALYZE → CLARIFY → FILE → IMPLEMENT │
├─────────────────────────────────────────────┤
│ │
│ 1. FETCH ISSUE │
│ gh issue view <number> --json ... │
│ → Get title, body, author, state │
│ │
│ 2. GET CONTRIBUTOR PROFILE │
│ gh api users/<login> │
│ → twitter_username, blog, bio, name │
│ → Store in semantic-memory for credits │
│ semantic-memory_store( │
│ information="Contributor @{login}: │
│ {name} (@{twitter} on Twitter). │
│ Filed issue #{number}. Bio: {bio}", │
│ tags="contributor,{login},issue-{#}" │
│ ) │
│ │
│ 3. ANALYZE │
│ → Is it a bug? Feature? Question? │
│ → Can you reproduce? │
│ → Is it in scope? │
│ │
│ 4. CLARIFY (if needed) │
│ → Ask for repro steps │
│ → Request context/versions │
│ → Genuine questions, not interrogation │
│ │
│ 5. FILE CELL │
│ hive_create( │
│ title="Issue #N: <summary>", │
│ type="bug|feature", │
│ description="<link + contributor>" │
│ ) │
│ │
│ 6. TAG ISSUE │
│ gh issue edit <number> --add-label bug │
│ │
│ 7. IMPLEMENT │
│ → Fix the issue │
│ → Write tests │
│ → Close cell │
│ │
│ 8. CREDIT IN CHANGESET │
│ → Add "Thanks @twitter" or │
│ "Thanks <name> (<blog>)" │
│ │
└─────────────────────────────────────────────┘
| Issue Type | Action | Create Cell? | Credit? | |------------|--------|--------------|---------| | Valid bug with repro | Confirm → file cell → fix | ✅ Yes | ✅ Yes | | Bug missing repro | Ask for steps → wait | ⏸️ Defer | ✅ Yes (when fixed) | | Feature request in scope | Validate → file cell → implement | ✅ Yes | ✅ Yes | | Feature out of scope | Explain why → close | ❌ No | ❌ No | | Duplicate | Link to original → close | ❌ No | ✅ Maybe (if original gets fixed) | | Question/support | Answer → close | ❌ No | ❌ No | | Already fixed | Confirm → close | ❌ No | ✅ Yes (if recent) |
# Get issue details
bun run scripts/issue-summary.ts <owner/repo> <number>
# Returns: title, body, author, state, labels, url
# Get contributor profile (includes Twitter!)
bun run scripts/get-contributor.ts <login> [issue-number]
# Example: bun run scripts/get-contributor.ts justBCheung 42
# Returns:
# - Profile details (name, twitter_username, blog, bio, avatar_url)
# - Ready-to-paste changeset credit: "Thanks to Brian Cheung ([@justBCheung]...)"
# - Ready-to-paste semantic-memory_store command
import { getIssueSummary } from "./scripts/issue-summary.ts";
import { getContributor } from "./scripts/get-contributor.ts";
// 1. Fetch issue
const issue = await getIssueSummary("owner/repo", 42);
// 2. Get contributor profile
const contributor = await getContributor(issue.author.login);
// 3. Store contributor in semantic-memory for future credits
semantic-memory_store({
information: `Contributor @${contributor.login}: ${contributor.name || contributor.login} ${contributor.twitter_username ? `(@${contributor.twitter_username} on Twitter)` : ''}. Filed issue #42. Bio: '${contributor.bio || 'N/A'}'`,
tags: `contributor,${contributor.login},issue-42`
});
// 4. Analyze and decide
if (issue.body.includes("TypeError") && issue.body.includes("steps to reproduce")) {
// Valid bug with repro - file cell
await hive_create({
title: `Issue #42: ${issue.title}`,
type: "bug",
description: `${issue.url}\n\nReported by: ${contributor.name || contributor.login}\nTwitter: ${contributor.twitter_username || 'N/A'}\n\n${issue.body.slice(0, 500)}`
});
// Tag issue
await $`gh issue edit 42 --add-label bug`;
} else if (!issue.body.includes("steps to reproduce")) {
// Missing info - ask nicely
await $`gh issue comment 42 --body "Hey ${contributor.name || contributor.login}! Could you share steps to reproduce? That'll help me track this down."`;
}
After filing cell:
Hey [name]! Thanks for reporting this. I've filed a tracking issue - we'll get this sorted.
After asking for clarification:
Hey [name], could you share [X]? That'll help me nail down what's happening.
After fixing:
Fixed in [commit]! Should be in the next release. Thanks for catching this 🙏
When closing as duplicate:
This is a dupe of #[N] - tracking there. Thanks for the report!
When closing as not-a-bug:
This is actually expected behavior because [reason]. If you're trying to [X], here's how: [link/example]
With name AND Twitter handle (PREFERRED):
---
"package-name": patch
---
Fixed [bug description]
Thanks to [Name] ([@twitter_username](https://x.com/twitter_username)) for the report!
With Twitter handle only (no name):
---
"package-name": patch
---
Fixed [bug description]
Thanks to [@twitter_username](https://x.com/twitter_username) for the report!
With name only (no Twitter):
---
"package-name": patch
---
Fixed [bug description]
Thanks to [Name] (@github_username on GitHub) for the report!
GitHub username only (no name, no Twitter):
---
"package-name": patch
---
Fixed [bug description]
Thanks to @github_username for the report!
Why include both name and Twitter? Names are human, Twitter handles enable engagement. "Thanks to Brian Cheung (@justBCheung)" gives credit AND makes it easy to tag them when tweeting the release.
GitHub user profiles have these useful fields:
{
"login": "bcheung",
"name": "Brandon Cheung",
"twitter_username": "justBCheung", // ← THIS!
"blog": "https://example.com",
"bio": "Building cool stuff",
"avatar_url": "...",
"html_url": "..."
}
Always fetch the profile - it's one API call and gives you credit info for changesets that get tweeted.
DO:
DON'T:
Examples:
❌ Corporate: "Thank you for your contribution. We have logged this issue and will investigate."
✅ Joel: "Hey Brandon! Thanks for catching this. I can reproduce it - looks like the auth refresh logic is borked. Tracking in #42."
❌ Interrogative: "Can you please provide the following information: 1) Version 2) Steps to reproduce 3) Expected behavior 4) Actual behavior"
✅ Joel: "Hey! Could you share which version you're on? And if you've got repro steps that'd be 🔥"
❌ Over-promise: "We'll fix this in the next patch release!"
✅ Joel: "On it! Should have a fix soon."
// File cell with issue reference
hive_create({
title: `Issue #42: Token refresh fails`,
type: "bug",
description: `https://github.com/owner/repo/issues/42
Reported by: Brandon Cheung
Twitter: @justBCheung
GitHub: @bcheung
User reports auth tokens aren't refreshing. Repro steps in issue.`
});
// When closing cell, reference in commit
git commit -m "fix: token refresh race condition
Fixes #42 - adds 5min buffer before token expiry.
Thanks @justBCheung for the report!"
scripts/get-contributor.ts - GitHub user profile fetcherscripts/issue-summary.ts - Issue details with smart formattinggh issue view, gh api users/<login>development
Patterns for testing code effectively. Use when breaking dependencies for testability, adding tests to existing code, understanding unfamiliar code through characterization tests, or deciding how to structure tests. Covers seams, dependency injection, test doubles, and safe refactoring techniques from Michael Feathers.
tools
Principles for building reusable coding systems. Use when designing modules, APIs, CLIs, or any code meant to be used by others. Based on "A Philosophy of Software Design" by John Ousterhout. Covers deep modules, complexity management, and design red flags.
development
Multi-agent coordination patterns for OpenCode swarm workflows. Use when working on complex tasks that benefit from parallelization, when coordinating multiple agents, or when managing task decomposition. Do NOT use for simple single-agent tasks.
development
Meta-skill for generating new skills with proper format and structure. Use when creating new skills for the swarm system or when agents need to generate skill scaffolds. Ensures skills follow conventions (frontmatter format, directory structure, bundled resources).