视频助手
- 作者仓库星标 39
- 作者更新于 实时读取
- 作者仓库 awesome-omni-skill
- 领域
- 通用
- 兼容 Agent
-
- Claude Code
- Cursor
- Cline
- Codex
- Windsurf
- Gemini CLI
- +20
- 信任分
- 88 / 100 · 社区维护
- 作者 / 版本 / 许可
- @diegosouzapw · 未声明 license
- Token 消耗评级
- 低消耗
- 接入复杂程度
- 即装即用
- 是否需要外部 API Key
- 不需要
- 兼容的系统
- 未声明(默认跨平台)
- 底层运行要求
- 无特殊要求
- 文件与系统权限
-
- 只读
- 允许写入 / 修改
- 网络行为
- 仅限本地
- 安装命令数
- 26 条
档案由构建时根据 SKILL.md 与安装命令自动衍生,可能与作者实际意图存在差异。
需要注意: 未限定 allowed-tools,默认拥有全部工具权限。
---
name: xradio
description: CLI tool for pulling, caching, and reading Twitter/X posts offline A command-line tool for manag…
category: 通用
runtime: 无特殊运行时
---
# xradio 输出预览
## PART A: 任务判断
- 适用问题:通用任务拆解、检查和交付。
- 输入要求:目标材料、限制条件、期望输出和验收方式。
- 证据边界:围绕“Prerequisites / Commands / Pull Tweets from Feed”读取原文规则,不把推断写成作者承诺。
## PART B: 执行结果
- **01** 任务判断:确认你的需求是否属于通用任务拆解、检查和交付,并标出输入、限制和预期结果。
- **02** 执行计划:优先按“Prerequisites / Commands / Pull Tweets from Feed”拆成步骤,说明每一步会读取什么、修改什么、产出什么。
- **03** 交付结果:给出可复制的命令、文件改动、检查清单或内容草稿,并说明如何继续迭代。
- **04** 风险边界:结合 读取文件、写入/修改文件、主要在本地完成、通常不需要额外 API Key 给出执行前确认项。
## Running Rules
- 读取文件、写入/修改文件;主要在本地完成;通常不需要额外 API Key。
- 先小样例验证,再放大到真实任务。
- 交付时同时给结果、检查口径和下一步迭代建议。 原文没有稳定的斜杠命令要求。安装验证后通常全局生效,直接在对话里点名这个 Skill 并描述任务即可。
告诉 Agent 目标文件或材料、期望结果、不可改范围、是否允许联网或执行命令。本 Skill 的权限画像是:读取文件、写入/修改文件。
先用一个小任务确认它会围绕“Prerequisites / Commands / Pull Tweets from Feed”工作;涉及文件或命令时,先看 diff、日志、预览或测试结果。
检查最终产物是否包含明确结果、必要证据和下一步动作;如果输出泛泛而谈,就补充输入、边界和验收标准后重跑。
---
name: xradio
description: CLI tool for pulling, caching, and reading Twitter/X posts offline A command-line tool for manag…
category: 通用
source: diegosouzapw/awesome-omni-skill
---
# xradio
## 什么时候使用
- 把通用方向的常用动作沉淀成 Agent 可调用的技能 适合处理通用任务拆解、检查、交付和复盘,核心价值是把输入、判断、执行、验证和交付边界固定下来,避免 Agent 泛泛回答。 把任务拆成可执行、可检查、可继续迭代的步骤;通常不需要额外…
- 面向通用任务拆解、检查和交付,优先处理能明确输入、步骤和验收标准的工作。
## 需要提供什么
- 目标材料、目录范围、期望结果和不可改动内容。
- 是否允许联网、执行命令、读写文件或调用外部服务。
## 执行规则
- 围绕「Prerequisites / Commands / Pull Tweets from Feed」组织步骤,不把推断写成作者事实。
- 读取文件、写入/修改文件;主要在本地完成;通常不需要额外 API Key。
- 先跑小样例,确认结果可检查后再扩大任务范围。
## 输出要求
- 给出最终产物、关键证据、验证方式和下一步动作。
- 信息不足时标记 unknown,不编造命令、平台或依赖。 作者原文负责流程事实;仓库文件负责来源和命令;流狐只补充适用场景、限制和质量判断。
skill "xradio" {
输入层 -> 用户目标 + 目标文件 + 禁止范围 + 验收标准
上下文层 -> Prerequisites / Commands / Pull Tweets from Feed
规则层 -> SKILL.md 触发条件 / 执行顺序 / 输出格式
运行层 -> 无特殊运行时 | 读取文件、写入/修改文件 | 主要在本地完成
安全层 -> 通常不需要额外 API Key + 小任务验证 + diff / 日志复核
输出层 -> 可复制结果 + 检查清单 + 下一步迭代
} xradio
A command-line tool for managing a local cache of Twitter/X posts. Pull tweets from your feed, read them offline, track read/unread status, and play videos.
Prerequisites
- Browser server must be running for
pullcommand:browser server start - Must be logged into Twitter/X in the browser session
yt-dlpandmpvfor video playback
Commands
Pull Tweets from Feed
xradio pull --limit <n>
Fetches tweets from your Twitter home feed and saves them to local storage.
--limit <n>: Number of tweets to pull (default: 10)- Automatically scrolls to load more if needed
- Never overwrites existing cached tweets - skips duplicates
- Requires browser server to be running and logged into Twitter
Example:
xradio pull --limit 20
Output:
📡 Pulling up to 20 new tweets from feed...
🌐 Browser connected: https://x.com/home
🔄 Refreshing feed...
scroll 1: +8 new, 2 cached (8/20 total new)
scroll 2: +6 new, 4 cached (14/20 total new)
scroll 3: +6 new, 3 cached (20/20 total new)
✅ [a1b2c3d] User Name: Tweet text preview...
✅ [e4f5g6h] Another User: More tweet content here...
📦 Done! Saved 20 new tweets.
List Cached Tweets
xradio list # Show unread tweets only
xradio list -A # Show all tweets (including read)
xradio list --all # Same as -A
Lists tweets in the offline cache.
- By default, only shows unread tweets
- Use
-Aor--allto include read tweets - Shows media indicators: 🎬 (video), 🖼️ (image)
- Shows ✓ for read tweets (when using --all)
Output:
📚 4 unread tweets in cache:
a1b2c3d🎬🖼️ Man, Crimson Desert looks like the kind of game that would make me disappear...
e4f5g6h🖼️ Breaking: New development in the ongoing situation as officials announce...
Use 'xradio read <id>' to view a tweet, 'xradio mark <id>' to mark as read.
Read a Tweet
xradio read <id>
Display the full content of a cached tweet.
- Supports short IDs (like git):
xradio read a1binstead of full hash - Shows formatted tweet with author, text, media indicators
- If ambiguous (multiple matches), asks for longer prefix
Example:
xradio read a1b
Output:
────────────────────────────────────────────────────────────
# User Name
**@handle** · 2h
This is the full tweet text content here...
---
🎬 Video | 🖼️ Image
[View on Twitter](https://x.com/handle/status/123456789)
────────────────────────────────────────────────────────────
ID: a1b2c3d (a1b2c3d4e5f6g7h8i9j0...)
Pulled: 2026-01-08T16:00:00.000Z
🎬 This tweet has a video. Use 'xradio play a1b' to play it.
Mark as Read
xradio mark <id>
Mark a tweet as read. It will no longer appear in xradio list (only in xradio list -A).
Example:
xradio mark a1b
# Output: ✅ Marked a1b2c3d as read
Unmark (Mark as Unread)
xradio unmark <id>
Remove the read status from a tweet, making it appear in the unread list again.
Example:
xradio unmark a1b
# Output: ↩️ Unmarked a1b2c3d (now unread)
Describe Images
xradio describe <id>
xradio describe <id> "custom prompt here"
Describe images in a tweet using AI vision. Downloads images and generates a text description suitable for radio narration.
- Works only on tweets with images (🖼️ indicator)
- Uses AI vision model to describe the image content
- Optional custom prompt for specific questions about the image
Example:
xradio describe 836
Output:
🖼️ Describing 1 image(s) from: Visegrád 24
This is a line graph from Netblocks.org showing normalized network
connectivity in Iran from January 7 to 8, 2026...
Play Video
xradio play <id>
Play a video from a tweet that has video content.
- Extracts video URL via yt-dlp
- Plays with mpv (video + audio)
- Interactive controls:
qto quit,spaceto pause
Example:
xradio play a1b
Output:
🎬 Extracting video URL for: User Name
https://x.com/handle/status/123456789
📺 Playing video...
(Press 'q' to quit, 'space' to pause)
Storage Format
Tweets are stored in storage/<id>.md where id is SHA1 hash of the permalink.
Each file has YAML frontmatter + Markdown content:
---
id: a1b2c3d4e5f6...
permalink: https://x.com/handle/status/123456789
displayName: User Name
handle: "@handle"
timestamp: "2h"
hasVideo: true
hasImage: false
pulledAt: 2026-01-08T16:00:00.000Z
offline:
read: true
readAt: 2026-01-08T17:00:00.000Z
---
# User Name
**@handle** · 2h
Tweet text content...
---
🎬 Video
[View on Twitter](https://x.com/handle/status/123456789)
Typical Workflow
# 1. Start browser (if not running)
browser server start
# 2. Pull fresh tweets
xradio pull --limit 10
# 3. See what's new
xradio list
# 4. Read a tweet
xradio read a1b
# 5. If it has video, play it
xradio play a1b
# 6. Mark as read when done
xradio mark a1b
# 7. Continue with next unread tweet
xradio list
Notes
- Short IDs work like git refs - use just enough characters to be unique
- The
offline.*namespace in frontmatter is for our metadata (read status, etc.) - Tweet data is never overwritten once cached - pull is additive only
- Video playback requires working yt-dlp Twitter extraction (may have intermittent issues with Twitter API)
先判断是否适合
作者设计意图
作者的方法与取舍
边界和复核