Getting it into your agent
This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.
/plugin marketplace add ggzggz0608/btc-news-radar/plugin install btc-news-radarWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/ggzggz0608/btc-news-radar/btc-radar)<a href="https://agentmods.dev/skills/ggzggz0608/btc-news-radar/btc-radar"><img src="https://agentmods.dev/badge/skills/ggzggz0608/btc-news-radar/btc-radar/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/ggzggz0608/btc-news-radar/btc-radar"><img src="https://agentmods.dev/badge/skills/ggzggz0608/btc-news-radar/btc-radar.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00179 | $0.02114 |
| Opus 5 | $0.00089 | $0.01057 |
| Sonnet 5 | $0.00036 | $0.00423 |
| Haiku 4.5 | $0.00018 | $0.00211 |
Grade A, and why
btc-radar scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 153 lines — stays where its author put it; the contents beside it link to each section on GitHub.
₿ BTC News Radar
一个本地比特币情报终端。数据、数据库、配置全部留在用户自己的机器上， 不上传任何地方。
你（Claude）该怎么用这个技能
所有操作都通过同一个零依赖脚本完成：
node "${CLAUDE_PLUGIN_ROOT}/skills/btc-radar/scripts/radar.mjs" <子命令>
| 子命令 | 什么时候用 |
|---|---|
setup |
首次安装，或用户说"装坏了/修一下" |
start |
启动采集与看板（后台常驻） |
stop |
停止 |
restart |
改完配置后 |
status |
用户问"跑起来没""有数据吗" |
doctor |
用户说"不工作""没数据""收不到推送" |
update |
用户要新版本 |
where |
需要知道安装路径时 |
autostart on|off|status |
用户要开机自启，或问"重启后还在吗" |
ensure |
计划任务内部调用，幂等，不用手动跑 |
执行完把关键输出转述给用户，不要只说"已完成"。 用户装这个是为了看数据， 他需要知道现在有多少条事件、评分后端通没通、看板地址是什么。
三条不能违反的边界
1. 不输出买卖建议
这个产品刻意不做"该做多还是做空"。原因有两条，转述给用户时两条都要说：
- 不是持牌投顾，不提供个性化投资建议
- 更重要：没有任何系统能可靠给出这种信号，声称能做到的工具长期只会让用户亏钱
它给的是市场结构描述 —— 事件影响分、确定性阶梯、技术面状态、蜡烛形态的几何与
位置。Key Zone 不是"入场点"，失效位 不是"止损"，偏多 不是"做多"。
用户问"我该买吗"，如实说这个工具不回答这个问题，然后把它能给的事实摆出来。
2. 会消耗用户自己的 Claude 订阅额度
AI 评分是通过本机的 Claude Code CLI（claude -p）跑的。这意味着：
- 不需要 API key，走用户已有的订阅
- 但消耗的是用户自己的额度，而且是持续消耗（默认每 20 秒最多处理 5 条）
- 实测一天几十到几百条事件
首次安装时必须主动告诉用户这件事，不要等他自己发现。
嫌多可以改 config/app.yaml 里的评分频率。
3. 不替用户做外发动作
手机推送要用户自己在手机上装 ntfy App 并扫码订阅 —— 这一步无法代劳。
不要把用户的 NTFY_TOPIC 贴到任何公开地方：那串就是唯一凭证，
任何知道它的人都能收到他的推送。
开机自启
autostart on 注册 Windows 计划任务：登录后 30 秒启动 + 每 5 分钟确保在运行。
开启前必须告诉用户：采集器会跟着自启，开机后就持续消耗他自己的 Claude / ChatGPT 额度。这一条不能省略。
只实现了 Windows。非 Windows 会明确报错 —— 不要假装能用。
平台支持
只在 Windows 上验证过。 macOS / Linux 的代码路径写了但没实际跑过 —— 安装路径推导与进程启停未经验证。
用户在 macOS / Linux 上遇到问题时，不要假装它应该能用：如实说明这部分没测过， 把报错原样收集起来，建议他去仓库开 issue。桌面通知在非 Windows 平台会自动跳过， 这是预期行为，手机推送不受影响。
安装位置
应用装在用户数据目录，不在插件目录里：
| 系统 | 路径 |
|---|---|
| Windows | %LOCALAPPDATA%\btc-news-radar |
| macOS | ~/Library/Application Support/btc-news-radar |
| Linux | $XDG_DATA_HOME/btc-news-radar（默认 ~/.local/share/…） |
原因：插件缓存目录带版本号，升级时整个被替换 —— 数据库放里面会在某次更新后消失。
用户想换位置就设 RADAR_HOME 环境变量。
环境要求
| 项 | 要求 | 缺了会怎样 |
|---|---|---|
| Node.js | ≥ 22.5 | 装不上。存储层用 node:sqlite，22.5 才有 |
| git | 任意版本 | 下载不了应用 |
| pnpm | 任意 | setup 会尝试用 corepack 自动启用 |
| Claude Code CLI | 已登录 | 采集照常跑，但评分会一直排队。登录后自动补评，不用重启 |
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 5d ago Changed · +11 lines 5b090343c888
- 12d ago First seen · 142 lines · 179 tokens per session scan A 7cd0f91ce53c
btc-radar is a skill published in the GitHub repository ggzggz0608/btc-news-radar (3 stars, last pushed 6d ago), licensed MIT. It adds 179 tokens to every session and 2,114 once invoked, about $0.0009 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
pricing-refresh
Refresh or verify the hermit cost rate table against Anthropic's live pricing. Use when the operator says "refresh pricing", "verify pricing", "is the rate table current", "new model pricing", or asks whether scripts/lib/pricing.ts matches current Codex rates. Maintainer-only; never autonomous, never a routine, never…
monitor
Autonomous monitoring loop. Checks open trades against SL/TP levels, closes trades that hit targets, evaluates expired predictions, and generates periodic summaries. Run via cron for full autonomy. Usage: /monitor.
validate-predictions
Review and validate pending predictions against current market data. Usage: /validate-predictions.
analyze
Run a comprehensive multi-agent crypto analysis with phased execution. Usage: /analyze BTC or /analyze ETH SOL.
invinoveritas
Pay-per-call agent tools over Bitcoin Lightning / USDC (x402), exposed as a remote MCP server. Use BEFORE any irreversible or consequential action (a trade, a destructive command, shipping code, spending funds) to get a capital-scale-aware governance review; for facts-only crypto market intelligence (macro risk…
close-trade
Close an open trade and run post-mortem analysis. Usage: /close-trade trade001 or /close-trade trade001 at 98500.