Jamailar/Beav

---

name: content-topic-miner

---

name: cosyvoice-ssml

---

name: image-director

---

name: image-prompt-optimizer

---

name: member-skill-distiller

description: Distill team members from profile, files, and YouTube subtitles into session-activated member skills.

---

name: redclaw-style-definition

description: RedClaw 空间首次创作前的自媒体定位与写作风格定义。用“账号首页截图 + 文章/对标内容截图”两步证据流完成定位和文风学习。

---

name: skill-creator

description: Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Codex's capabilities with specialized knowledge, workflows, or tool integrations.

---

name: social-cover-director

description: Use when the user wants to plan or generate a social media cover image for a Xiaohongshu note, Douyin/TikTok photo post, Reels/Shorts cover, WeChat/Video Account post, Bilibili/Weibo post, Pinterest pin, Instagram cover, Facebook/Meta social post, or similar creator/social content. Handles platform-native cover routing, aspect ratio, scroll-stopping hook, exact on-image copy, reference-image roles, and image.generate prompt construction for one cover or a small variant set.

---

name: tts-director

---

name: video-director

description: Canonical entrypoint for every AI chat request that asks to make, generate, plan, or edit a video, including promotional films, ads, short videos, product videos, reference-image videos, image-to-video, first/last-frame transitions, storyboard/keyframes, character talking-head videos, 宣传片, 广告片, 短视频, 视频, 口播视频, 分镜图, 图片转视频, and 商品宣传片. Do not invoke cosyvoice-ssml or tts-director as the first response to a video request. This skill must draft a script and shot table, generate storyboard contact-sheet preview images through image.generate, ask the user to confirm the script plus storyboard together, and only then call video generation.

---

name: wander-synthesis

description: 漫步选题方法，用于从随机或编排器选出的素材中选出高互动母版、拆解其内容公式，再从小批量素材里挖掘细小选题。

---

name: writing-style

---

name: wwud

description: What Would User Do for RedConvert. Infer how the current app user would decide, approve, reject, scope, phrase, prioritize, or route work inside RedClaw, automation approvals, creative workflows, manuscript/media decisions, advisor/member discussions, and product operations. Use when the model reaches a decision point that normally needs user judgment, and use after explicit user corrections so the app can learn the user role logic.

---

name: xhs-comment-insight

---

name: xhs-title

description: 小红书标题公式工具。从 75 个验证过的爆款公式中，帮你挑对的、用对的、理解为什么用这个。 触发方式：/xhs-title、/小红书标题、「帮我起个小红书标题」「小红书标题公式」 Xiaohongshu title formula tool. Pick the right formula from 75 proven templates. Trigger: /xhs-title, "xiaohongshu title", "RED title formula"

---

name: app-feedback-admin

description: Read platform-admin feedback issues through the trusted HTTP admin API, summarize user reports, and turn them into RedBox/RedConvert repair and iteration plans. Use when the user asks to read app feedback, triage feedback tickets, inspect remote feedback context, or prepare fix/iteration plans from feedback. This skill must keep FEEDBACK_ADMIN_API_KEY server-side/local only and must not place it in app UI, frontend code, desktop client code, logs, commits, or shared artifacts.

---

name: content-topic-miner