[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-80223":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":9,"language":10,"languages":9,"totalLinesOfCode":9,"stars":11,"forks":12,"watchers":13,"openIssues":14,"contributorsCount":15,"subscribersCount":15,"size":15,"stars1d":14,"stars7d":16,"stars30d":17,"stars90d":15,"forks30d":15,"starsTrendScore":18,"compositeScore":19,"rankGlobal":9,"rankLanguage":9,"license":9,"archived":20,"fork":20,"defaultBranch":21,"hasWiki":22,"hasPages":20,"topics":23,"createdAt":9,"pushedAt":9,"updatedAt":24,"readmeContent":25,"aiSummary":26,"trendingCount":15,"starSnapshotCount":15,"syncStatus":27,"lastSyncTime":28,"discoverSource":29},80223,"claude-code-workflow-creator","ray-amjad\u002Fclaude-code-workflow-creator","ray-amjad","A Claude Code skill for authoring Workflow tool scripts: deterministic multi-agent orchestration. Preview of an unreleased feature.",null,"JavaScript",79,9,63,1,0,5,15,6,3,false,"main",true,[],"2026-06-12 02:03:59","# claude-code-workflow-creator\n\nA Claude Code **skill** that teaches Claude to author **workflows**: deterministic\nmulti-agent orchestration scripts that fan work out to fresh-context subagents\nunder plain JavaScript control flow.\n\nA workflow is a JavaScript file. The loops, the conditionals, the fan-out are\nordinary code that you control. Only the leaf `agent()` calls spend model tokens,\nand each one runs in its own clean context window. The result is multi-agent work\nthat behaves the same way every run and can be resumed if it stops partway.\n\nThis skill carries the file format, the judgement calls, and a tested authoring\nprocedure, so you can just ask Claude to \"create a workflow for X\" and get a\ncorrect, runnable file back.\n\n## Heads up: the Workflow tool is not released yet\n\nThis skill writes files for Claude Code's **Workflow tool**. As of right now that\ntool is **unreleased and unannounced**.\n\nIt ships inside the Claude Code binary, but it stays hidden behind an environment\nvariable, and Anthropic has not officially documented or announced it. That can\nchange at any time, including how it is enabled and how it behaves.\n\n**What this means for you:**\n\n- You can install this skill today and read through it to understand workflows.\n- The workflows it writes will only actually *run* once the Workflow tool is\n  switched on in a released version of Claude Code.\n- The right move is to wait for Anthropic to officially announce and release the\n  feature, then come back and use this skill once it is live.\n\nTreat this repo as a preview until then. When the feature ships, everything here\nis ready to go.\n\n## What is in this repo\n\nThe repo root is the skill itself. Drop it into your Claude Code skills folder\nand Claude picks it up automatically.\n\n| Path | What it is |\n|------|------------|\n| `SKILL.md` | The skill entry point: the procedure Claude follows to design and write a workflow. |\n| `references\u002Fapi-reference.md` | The complete manual: every global, every option, every cap and constant. |\n| `references\u002Fpatterns.md` | Copy-paste orchestration patterns (fan-out, pipeline, loop-until-budget, judge panel, and more). |\n| `assets\u002Ftemplates\u002F` | Starter files for the three core shapes: fan-out, pipeline, loop. |\n| `assets\u002Fexamples\u002F` | Six complete, runnable example workflows, with a README mapping each one to a technique. |\n| `scripts\u002Fvalidate-workflow.mjs` | A linter that checks a workflow file against the parser's hard rules before you run it. |\n\n## Install\n\nCopy the skill into your global Claude Code skills folder:\n\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002Fray-amjad\u002Fclaude-code-workflow-creator.git\nmkdir -p ~\u002F.claude\u002Fskills\ncp -R claude-code-workflow-creator ~\u002F.claude\u002Fskills\u002Fworkflow-creator\n```\n\nThat is all. The next time Claude Code starts, the skill is available. Ask Claude\nto \"create a workflow\" and it will trigger.\n\n## Using it (once the Workflow tool is released)\n\n1. Enable the Workflow tool. Today it is gated behind an environment variable:\n\n   ```bash\n   export CLAUDE_CODE_WORKFLOWS=1\n   claude\n   ```\n\n   When the feature is officially released, follow Anthropic's instructions for\n   turning it on, since the exact mechanism may change.\n\n2. Ask Claude to build one, for example: \"create a workflow that reviews my\n   branch across bugs, security, and tests, then verifies each finding.\"\n\n3. Claude uses this skill to write the file, validate it, and run it. Watch live\n   progress with the `\u002Fworkflows` command.\n\n## A note on accuracy\n\nThe details in this skill (the file format, the determinism rules, the caps, how\nto set a model, how structured output works) were checked directly against the\nClaude Code binary, not guessed. Because the feature is pre-release, some of it\nmay shift before launch. If something stops matching once the feature is\nofficially out, open an issue.\n\n## Credits\n\nBuilt and maintained by [Ray Amjad](https:\u002F\u002Fwww.youtube.com\u002F@RAmjad). If this is\nuseful, the channel covers more agentic coding workflows like this one.\n","该项目是一个用于编写工作流工具脚本的Claude Code技能，支持确定性的多代理编排。其核心功能是通过JavaScript控制流程来创建和管理多代理任务分发，使得每个代理在独立的上下文中运行，确保每次执行结果的一致性。项目提供了一套完整的文件格式、判断逻辑及经过测试的创作流程，便于用户请求Claude生成特定需求的工作流脚本。尽管目前所对应的工作流工具尚未正式发布，但该技能已准备好配合未来版本使用，适用于需要稳定且可恢复的自动化任务处理场景中。",2,"2026-06-11 03:59:44","CREATED_QUERY"]