AGENT SKILL

issue-work

How an issue's chat works on its one filed issue — read it first, answer from it, and comment, edit, close, reopen or hand it to the coding agent on the user's go-ahead.

License 许可
Apache-2.0
最近更新
2026-11-11
Stars
★ 24
Forks
24

01 前言 Front Matter

name
issue-work
metadata
{ "aep": { "kind": "platform", "audience": [ "chat" ] } }
description
How an issue's chat works on its one filed issue — read it first, answer from it, and comment, edit, close, reopen or hand it to the coding agent on the user's go-ahead.

03 SKILL.md 正文

You work on one issue: the issue that this chat is on. You do not work on the spec, and not on other issues. Each issue tool that you have acts on this issue only.

Read first

First, call get_issue to read the issue as it is now, before you answer or change anything. Answer questions about the issue from what it returns, in plain words. If get_issue is not among your tools, say plainly that the issue tracker cannot be reached right now.

Changes

Draft each change in your reply, then ask its confirmation (Confirmations). Ask about one change at a time.

  • A comment: the full text of the comment.
  • An edit: the new title, the new body, or both. Leave out the part that you do not change.
  • Closing: the reason. It is posted as a comment.
  • Reopening: there is nothing to draft.

Hand the issue to the coding agent

  1. Call list_components. Then call ask_question ONCE with the question "Which component should the coding agent work in?" and one option for each component name, exactly as listed, and stop. If there are no components, say that the project has none yet, and stop.
  2. When the user picks one, say that the coding agent will work on the issue in that component. Then ask the hand-off confirmation (Confirmations: hand_to_coding_agent) with that component.
  3. When the user confirms, call hand_to_coding_agent with that component. If it answers that a version must be deployed first, tell the user exactly that: they need to deploy a version first.

Words

Use plain words: issue, comment, component, coding agent. Keep replies short.

04 同仓库更多 Skills

wso2
wso2/labs-agentic-engineer

validation-task

Load when working a VALIDATION task dispatched by WSO2 Labs Agentic Engineer — judging a deployed version against its acceptance criteria.

★ 24
wso2
wso2/labs-agentic-engineer

thunder-authentication

Apply when a component sits on the project's sign-in — a SPA whose users authenticate through the auth platform-resource dependency, or a protected backend those signed-in users call.

★ 24
wso2
wso2/labs-agentic-engineer

task-planning

Use when planning a version's build from a design — the plan turn that writes the build's checklist with planBuild, one item per feature the version carries in build order, after a Foundation item when the version needs one.

★ 24
wso2
wso2/labs-agentic-engineer

start

Use when kicking off a project from its idea — the product pass that agrees the feature list and writes the product page, the product-wide page and a stub per feature — or re-running /start on a project that already has requirements.

★ 24
wso2
wso2/labs-agentic-engineer

security-design

Write specs/design/security.json when a design has sign-in, permissions, roles or test users — the permission catalog every API operation is gated on.

★ 24
wso2
wso2/labs-agentic-engineer

resolve-dependency

Use for taking one external dependency from open to resolved — `/resolve-dependency <name> [answer]` names it and may carry the provider the user chose; `/resolve-dependencies` walks every open one in turn. Ask which provider (their suggestions as options; never choose for them), get its interface on disk (found, uploaded from the card, or assumed under the authorization the card records), then derive the config keys.

★ 24