Menu
Validates and optimizes run_pipeline DOT graphs with model selection from QoS catalog
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Recursively crawl websites using headless Chrome. Triggers: crawl, scrape website, 爬取, crawl site, deep crawl, website content.
OminiX ASR (speech-to-text), preset-voice TTS with emotion/speed control, and model management via Qwen3 models on Apple Silicon. For voice cloning and custom voice profiles, use mofa-fm. Triggers: voice, transcribe audio, text to speech, speak this, read aloud, model management, download model, 语音识别, 语音合成, 模型管理.
Deep multi-round web research with parallel fetching. Triggers: deep search, research, 深度搜索, 调研, investigate, deep research.
Manage sub-accounts under the current profile. Triggers: create account, 创建账号, sub account, manage account, list accounts, 子账号.
Send emails via SMTP or Feishu/Lark Mail. Triggers: send email, 发邮件, email to, 发送邮件, mail, send mail.
Get current weather for any city worldwide. Triggers: weather, forecast, temperature, 天气, 气温, how cold, how hot, is it raining, wind.
| name | pipeline-guard |
| description | Validates and optimizes run_pipeline DOT graphs with model selection from QoS catalog |
| version | 0.1.0 |
| author | octos |
| always | true |
Automatically validates DOT graphs and assigns optimal models before run_pipeline executes.
This is a lifecycle hook — it runs transparently before every pipeline execution.
model_catalog.json scores and assigns:
The LLM writes DOT graphs without model= attributes. This hook injects them automatically based on live QoS scores.