supermatrix
supermatrix에는 wzser에서 수집한 skills 13개가 있으며, 저장소 수준 직업 범위와 사이트 내 skill 상세 페이지를 제공합니다.
이 저장소의 skills
Use when a Codex session must manage an existing hosted mailbox through IMAP or SMTP, check mailbox access from the current project directory path, archive inbound mail locally, or send mail from an approved mailbox.
Create Excalidraw diagram JSON files that make visual arguments. Use when the user wants to visualize workflows, architectures, or concepts.
Use when the user expresses intent to promote current conversation insights into the global Principles docs (console-principles / coding-principles / business-principles). Typical trigger phrases include "记到 principle 里"、"加到 principles"、"沉淀到原则"、"加到 coding principles"、"这个规则应该通用化"、"put this in principles"、"promote this to principles". Do NOT activate for questions about existing Principles content or unrelated conversations — only when the user explicitly wants to contribute NEW content to the Principles docs. Drops a request file into the first-principle session's requests queue for async review; does not edit Principles templates directly.
Use when Codex needs to upload a local image or file to S.EE and return a public URL plus upload metadata such as the page URL, delete URL, file ID, hash, and source path. Trigger this when the user asks for a public image link, wants to upload a local screenshot, or needs complete S.EE upload metadata rather than a bare URL.
Use when Codex needs to fetch Lingxing OpenAPI data, especially ABA downloads, ad report pulls, ASIN diagnostics, seller metadata, category rank probes, or Amazon listing updates for title, bullets, description, and image slots through the whitelisted server with stable signing, retry, and output handling.
Use when SuperMatrix needs to browse, search, download, upload, create directories for, or rename files in the company NAS material library.
Use when you have a written implementation plan (typically from superpowers:writing-plans) and want to execute it via SuperMatrix's 4-mode supervisor loop — inline / mixed subagents / all-Codex sub-sessions / all-Claude sub-sessions — with reverse-backend independent verification. Reuses YOLO's plan.json + allocation gate + L1/L2 verifier bins instead of reimplementing them. Skip this skill and use superpowers:executing-plans or superpowers:subagent-driven-development directly if you don't need cross-backend dispatch.
Activate ONLY when the user message contains the exact literal string "ping skill-probe". This skill exists solely to verify cross-backend skill discovery. When activated, identify your backend, report the canonical path, and stop. Do not activate on any other input.
Use when adding a new model endpoint to the local registry, choosing a managed model for an implementation task, or generating a direct connection plan with api_key, base_url, pricing, and network requirements.
Use when querying Amazon marketplace data — product research, keyword analysis, category market sizing, competitor tracking, TikTok e-commerce data, or 1688 sourcing. Also use for market trend analysis, product selection (选品), and cross-platform e-commerce intelligence.
Layered web browser access over managed Chrome profiles (CDP proxy), with soft routing to generic/amzlisting/amzh10 packs.
微信读书助手 — 搜索书籍、管理书架、查看笔记划线、浏览书评、阅读统计、发现推荐好书
Control Ziniao Browser via the local Ziniao bridge. On skill load or before first invoke, GET /zclaw/tools and treat returned name list as the only allowed tool strings; then POST /zclaw/tools/invoke. API key for invoke via ~/.zclaw/config.json or ZCLAW_API_KEY. On bridge failure stop the turn per skill.