本文へ移動
cccskills
無料GitHub で公開

workflow-creator

Create or update reusable Dynamic Workflow JavaScript files under .qwen/workflows. Use when the user asks to create, save, edit, or reuse a Dynamic Workflow, including requests started from the Web Shell Workflows page.

インストール方法を見る

含まれるファイル(2)

  • SKILL.md3.4 KB
  • SKILL.test.ts1.5 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Workflow Creator

Create and maintain saved Dynamic Workflows for the current workspace.

Boundary

  • This skill manages .qwen/workflows/<name>.js files used by the workflow tool and exposed as /<name> slash commands.
  • Do not create or edit qwen-workflow-design/*.yaml; those Task Flow definitions are a different feature.
  • Use project scope by default. Write to ~/.qwen/workflows only when the user explicitly asks for a workflow shared across projects.

Workflow

  1. Inspect the current task and any existing workflow with the requested name. Ask a question only when the goal, ordering, or write scope is materially ambiguous.
  2. Choose a lower-case name containing only letters, digits, and hyphens. It must start with a letter and be at most 41 characters.
  3. Create the smallest script that captures the requested phases, dependencies, and final result. Do not add speculative branches, retries, or agents.
  4. Read the saved file back and verify its name, metadata, phase order, dependency flow, and final return value. Do not execute it unless the user also asks to run it.
  5. Report the saved path and slash command. In the interactive TUI's ink renderer, the slash command runs in the foreground: progress stays in the tool card, and completion displays the result and notifies the model. The OpenTUI renderer does not yet run client-scheduled tools; there, ask the model to call Workflow({ name: '<name>' }) instead. /workflows <runId> shows the run after it settles. In Web Shell, tell the user to return to Workflows and refresh the Saved tab if it is already open.

Script contract

  • Start with a literal metadata declaration:
export const meta = {
  name: 'Release readiness',
  description: 'Inspect, validate, and summarize a release candidate',
};
  • Use the sandbox globals phase(title), log(message), agent(prompt, options?), parallel(thunks), pipeline(items, ...stages), workflow(nameOrRef, args?), args, and budget. The workflow-authoring skill is the full reference for them — load it before writing anything beyond a trivial script.
  • Scripts cannot import modules or access the filesystem, shell, environment, or network directly. Put required reads and actions in explicit agent prompts.
  • Give every agent a complete, scoped prompt and a concise label. State whether it may edit files.
  • Express real concurrency as parallel([() => agent(...), () => agent(...)]). Do not pass already-started promises to parallel.
  • Keep dependent work sequential and pass prior results explicitly.
  • Put variable user input in args instead of hard-coding one-off values.
  • End every successful path with an explicit return of the final result. A trailing expression is not a return value.
  • Do not use node --check for validation: valid workflow scripts may contain top-level await and return because the runtime wraps them in an async function.

Updates

  • Preserve unrelated behavior and metadata when editing an existing workflow.
  • Do not overwrite an existing workflow with a different design unless the user requested that update.
  • Do not delete or rename a workflow unless the user explicitly asks.

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Reference for briefing a subagent or a fork — what to put in the prompt, what not to delegate, and a worked example. Load before writing a delegation prompt; the Agent tool's own description carries the launch rules and the background-agent rules.

日本語の概要は準備中です。原文の説明を表示しています。

QwenLM/qwen-code2.8万2026年10月11日 更新

Use after a Codex or Claude Code feature has been implemented in Qwen Code to run the selected reference agent and Qwen Code under the same scenario, capture HTTP and terminal traces, compare request bodies, tool/function schemas, outputs, and iterate until the reproduced behavior is close enough.

日本語の概要は準備中です。原文の説明を表示しています。

QwenLM/qwen-code2.8万2026年10月11日 更新

Use when reproducing an existing Codex or Claude Code feature in Qwen Code or another agent CLI by choosing a reference agent, capturing HTTP request bodies, prompts, tool/function schemas, terminal output, and then implementing the matching behavior in the target repo.

日本語の概要は準備中です。原文の説明を表示しています。

QwenLM/qwen-code2.8万2026年10月11日 更新

autofix

無料

Review and repair current local changes until they converge, or run Qwen Code Autofix issue and review workflows from GitHub Actions.

日本語の概要は準備中です。原文の説明を表示しています。

QwenLM/qwen-code2.8万2026年10月11日 更新

batch

無料

Execute batch operations on multiple files in parallel. Automatically discovers files, splits into chunks, and processes with parallel worker agents. Use `/batch` followed by operation and file pattern. For many independent single-turn transforms (translate/rewrite/extract each file into a new file) that can wait minutes to hours, you may suggest the user type `/batch-api` themselves for the half-price asynchronous Batch API — you cannot invoke it, and it is not suited to in-place edits or tasks needing tool feedback.

日本語の概要は準備中です。原文の説明を表示しています。

QwenLM/qwen-code2.8万2026年10月11日 更新

batch-api

無料

Prepare a many-file, single-turn transform (translate, rewrite, extract) as a plan and submit it to the asynchronous, half-price DashScope Batch API; results are delivered as new files hours later. Invoke explicitly with /batch-api.

日本語の概要は準備中です。原文の説明を表示しています。

QwenLM/qwen-code2.8万2026年10月11日 更新

QwenLM のスキルをすべて見る

このスキルの問題を報告する