[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-79980":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":14,"subscribersCount":14,"size":14,"stars1d":15,"stars7d":16,"stars30d":17,"stars90d":14,"forks30d":14,"starsTrendScore":16,"compositeScore":18,"rankGlobal":9,"rankLanguage":9,"license":19,"archived":20,"fork":20,"defaultBranch":21,"hasWiki":22,"hasPages":20,"topics":23,"createdAt":9,"pushedAt":9,"updatedAt":24,"readmeContent":25,"aiSummary":26,"trendingCount":14,"starSnapshotCount":14,"syncStatus":15,"lastSyncTime":27,"discoverSource":28},79980,"comfyui-agent-skill","MieMieeeee\u002Fcomfyui-agent-skill","MieMieeeee","Enable agents to run registered ComfyUI workflows through a local ComfyUI server using this repo’s CLI, returning structured JSON outputs.",null,"Python",83,14,73,0,2,6,10,3.53,"Apache License 2.0",false,"main",true,[],"2026-06-12 02:03:56","# ComfyUI Agent Skill\n\nThis repository is an Agent Skill folder (Claude Code \u002F Claude.ai \u002F Agent Skills). The only required file is [SKILL.md](SKILL.md).\n\n- 中文文档: [README.zh-CN.md](README.zh-CN.md)\n\n## Status\n\n- Requires a local or trusted self-hosted ComfyUI server (this repo is not a hosted service).\n- Not a hosted generation service; this package does not provide a ComfyUI backend.\n- This package does not install ComfyUI itself.\n- Stable interface: registered workflows only + CLI with structured JSON output.\n- Recommended trust model: do not point this at an untrusted public ComfyUI endpoint.\n- PyPI package: `comfyui-agent-skill-mie` (includes `comfyui-skill` alias).\n\n## Registered Workflows\n\nStable (reviewed configs in `assets\u002Fworkflows\u002F*.config.json`):\n\n- `z_image_turbo` (text-to-image)\n- `klein_edit` (image edit)\n- `qwen3_tts` (text-to-speech)\n- `ltx_23_t2v_distill` (text-to-video)\n- `ltx_23_i2v_distilled` (image-to-video)\n- `ace_step_15_music` (music\u002Faudio)\n- `qwen_image_2512_4step` (text-to-image)\n\nSource of truth: the runtime registry is derived from `assets\u002Fworkflows\u002F*.config.json` (and the corresponding `assets\u002Fworkflows\u002F*.json` workflow files). If this list drifts, trust the configs and `python -m comfyui generate --help` output.\n\n## For Skill Users (Generate \u002F Edit \u002F Video \u002F Audio)\n\nUse this when you want to run registered ComfyUI workflows via the CLI in this repo.\n\n- Primary entry: [SKILL.md](SKILL.md)\n- Workflow selection and sizing: [references\u002Fworkflows.md](references\u002Fworkflows.md)\n- Full CLI contract (output paths, async submit\u002Fpoll, JSON schemas, error codes): [references\u002Fcli.md](references\u002Fcli.md)\n- Prompt enhancement playbooks: [references\u002Fprompt_enhancement\u002F](references\u002Fprompt_enhancement\u002F)\n\n## Quick Start\n\n### Install (Recommended: pipx)\n\n```bash\npipx install comfyui-agent-skill-mie\ncomfyui-agent-skill-mie check\ncomfyui-agent-skill-mie generate -p \"a cute cat sitting on a windowsill at golden hour\"\ncomfyui-skill check\ncomfyui-skill generate -p \"a cute cat sitting on a windowsill at golden hour\"\n```\n\n### Alternative: uv tool install\n\n```bash\nuv tool install comfyui-agent-skill-mie\ncomfyui-agent-skill-mie check\ncomfyui-agent-skill-mie generate -p \"a cute cat sitting on a windowsill at golden hour\"\ncomfyui-skill check\ncomfyui-skill generate -p \"a cute cat sitting on a windowsill at golden hour\"\n```\n\n### Source mode (for development \u002F maintainers)\n\n```bash\nuv sync\nuv run --no-sync python -m comfyui check\nuv run --no-sync python -m comfyui generate -p \"a cute cat sitting on a windowsill at golden hour\"\n```\n\nOptional short command (after `uv sync` installs the project):\n\n```bash\nuv run --no-sync comfyui-skill check\nuv run --no-sync comfyui-skill generate -p \"a cute cat sitting on a windowsill at golden hour\"\n```\n\nFrom a local wheel (for testing before PyPI publish):\n\n```bash\npipx install dist\u002Fcomfyui_agent_skill_mie-*.whl\n```\n\nOr install from GitHub:\n\n```bash\npipx install \"git+https:\u002F\u002Fgithub.com\u002FMieMieeeee\u002Fcomfyui-agent-skill.git\"\n```\n\nIn tool-install mode, workflows\u002Freferences are read from the installed package, while writable data goes to a per-user directory:\n\n- Windows: `%APPDATA%\\\\comfyui-skill`\n- macOS: `~\u002FLibrary\u002FApplication Support\u002Fcomfyui-skill`\n- Linux: `$XDG_DATA_HOME\u002Fcomfyui-skill` or `~\u002F.local\u002Fshare\u002Fcomfyui-skill`\n\nShort alias also available: `comfyui-skill`\n\n## End-to-End Example (User input → decision → workflow → result)\n\nThis project is designed for agent-first workflow selection:\n\n- The Agent makes the primary workflow decision using [SKILL.md](SKILL.md), workflow capability metadata, and [references\u002Fworkflows.md](references\u002Fworkflows.md).\n- The CLI only provides a low-risk fallback when no explicit `--workflow` is chosen.\n\n### 1) User input (natural language)\n\n> \"帮我做一张海报，标题写「夏日促销」，风格现代简洁，主色调蓝白。\"\n\n### 2) Decision (what gets chosen and why)\n\n- Task type: text-to-image\n- Strong cue: poster \u002F embedded text requirement (\"海报\", \"标题写…\")\n- Prefer a stronger match over the general fallback T2I workflow\n- Selected workflow: `qwen_image_2512_4step`\n- Size choice: `704x1280` (HD portrait preset for posters)\n\n### 3) CLI command (agent explicitly chooses the workflow)\n\nIn this example, the agent explicitly chooses `qwen_image_2512_4step`; the CLI fallback selector is only a convenience when no explicit workflow is passed.\n\n```bash\ncomfyui-skill generate \\\n  --workflow qwen_image_2512_4step \\\n  --width 704 --height 1280 \\\n  -p \"Modern minimalist marketing poster, blue and white palette, bold Chinese title text '夏日促销', clean layout, high contrast, product-focused composition\"\n```\n\n### 4) Example JSON result (stdout)\n\n```json\n{\n  \"success\": true,\n  \"workflow_id\": \"qwen_image_2512_4step\",\n  \"status\": \"completed\",\n  \"outputs\": [\n    {\n      \"path\": \"C:\\\\\\\\Users\\\\\\\\\u003Cyou>\\\\\\\\AppData\\\\\\\\Roaming\\\\\\\\comfyui-skill\\\\\\\\results\\\\\\\\20260504\\\\\\\\120301_\u003Cjob_id>\\\\\\\\output.png\",\n      \"filename\": \"output.png\",\n      \"size_bytes\": 1234567\n    }\n  ],\n  \"error\": null,\n  \"job_id\": \"\u003Cprompt_id>\",\n  \"metadata\": {\n    \"seed\": 123456789,\n    \"prompt_id\": \"\u003Cprompt_id>\",\n    \"prompt\": \"Modern minimalist marketing poster ...\",\n    \"width\": 704,\n    \"height\": 1280\n  }\n}\n```\n\n## Adding Your Own Workflow (User Extension Example)\n\nIf you want to add your own ComfyUI workflow and make it available to the skill, treat it as “connecting a proven ComfyUI graph as a reviewed capability”.\n\n### 1) Export a working workflow from ComfyUI\n\nFirst make sure the workflow runs inside ComfyUI, then export it as API-format JSON.\n\n### 2) Import into your per-user workflow registry\n\nRun:\n\n```powershell\npython -m comfyui import-workflow .\\my_workflow.json --id my_custom_workflow\n```\n\nThis writes files under your per-user skill data directory (not the packaged built-in workflow registry), for example:\n\n```text\n\u003Cuser_data_root>\u002Fcustom_workflows\u002Fmy_custom_workflow\u002F\n  workflow.json\n  workflow.source.json\n  workflow.config.template.json\n```\n\n### 3) Review and activate the config\n\n- Open `workflow.config.template.json`\n- Confirm the minimal required fields:\n  - `description`, `capability`, `output_kind`, and the exposed inputs in `node_mapping`\n- Optional: improve agent choice by adding selection metadata:\n  - `intent_categories`, `input_modes`, `priority`, `keywords_any`, `selection_guidance`\n- Rename the reviewed file to:\n\n```text\nworkflow.config.json\n```\n\nOnly `workflow.config.json` is treated as an activated registered workflow. The template is not loaded.\n\nUser-added workflows are stored in the per-user skill data directory, not in the packaged built-in workflow registry, so upgrading the skill will not overwrite them.\n\n### 4) Run using your custom workflow\n\n```powershell\ncomfyui-skill generate --workflow my_custom_workflow -p \"test prompt\"\n```\n\nIf you omit `--workflow`, the CLI may apply low-risk fallback selection (for example, poster-like keywords) but it will not attempt to fully “understand” arbitrary user workflows. Agent-first selection remains the intended model.\n\n## Upgrade\n\nIf you installed via pipx:\n\n```bash\npipx upgrade comfyui-agent-skill-mie\n```\n\nIf you installed via uv tool:\n\n```bash\nuv tool upgrade comfyui-agent-skill-mie\n```\n\n## Troubleshooting\n\n- Run an environment doctor check (server + workflow preflight):\n  - `comfyui-skill doctor` (recommended)\n  - `comfyui-agent-skill-mie doctor`\n  - `uv run --no-sync python -m comfyui doctor` (source mode)\n- If the agent\u002Fskill runs inside WSL\u002Fcontainer\u002Fsandbox while ComfyUI runs on the host OS, `127.0.0.1` may point to the runtime itself instead of the host. Try `--server http:\u002F\u002Flocalhost:8188` or the host machine IP (and optionally persist it via `save-server`).\n- `SERVER_UNAVAILABLE`: ComfyUI is not reachable at the target URL. Start ComfyUI or re-run with `--server http:\u002F\u002F\u003Cip>:8188`.\n- `PREFLIGHT_MISSING_NODES`: install\u002Fenable required custom nodes on the ComfyUI server.\n- `PREFLIGHT_MISSING_MODELS`: download required model files on the ComfyUI server.\n- `NO_OUTPUT`: workflow ran but no media could be retrieved; check the workflow output node and server logs\u002FUI.\n- For `PREFLIGHT_MISSING_NODES`, `PREFLIGHT_MISSING_MODELS`, or `NO_OUTPUT`, consult the dependency reference: [references\u002Fworkflow_nodes.md](references\u002Fworkflow_nodes.md).\n\n## For Maintainers (Add \u002F Review Workflows)\n\nMaintenance docs are intentionally kept out of `SKILL.md` to keep the skill instructions user-focused.\n\n- Maintainer entry: [MAINTAINER.md](MAINTAINER.md)\n- Detailed workflow registration guide: [references\u002Fextension.md](references\u002Fextension.md)\n","该项目通过CLI使代理能够运行注册的ComfyUI工作流，并通过本地ComfyUI服务器返回结构化的JSON输出。其核心功能包括支持多种类型的工作流，如文本到图像、图像编辑、文本到语音、文本到视频等，使用Python编写，提供稳定的接口和结构化的输出。技术特点上，它需要用户自备或信任的自托管ComfyUI服务端，不直接安装ComfyUI本身。适用于需要自动化处理多媒体内容生成与编辑任务的场景，比如创意设计、内容制作等领域。","2026-06-11 03:58:46","CREATED_QUERY"]