Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add ChanningLua/prax-agent --skill hotspot-articlegit clone --depth 1 https://github.com/ChanningLua/prax-agentWrote 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/channinglua/prax-agent/hotspot-article)<a href="https://agentmods.dev/skills/channinglua/prax-agent/hotspot-article"><img src="https://agentmods.dev/badge/skills/channinglua/prax-agent/hotspot-article/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/channinglua/prax-agent/hotspot-article"><img src="https://agentmods.dev/badge/skills/channinglua/prax-agent/hotspot-article.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 | $0.00039 | $0.05548 |
| Opus 5.5 | $0.00016 | $0.02219 |
| Sonnet 5.5 | $0.00008 | $0.01110 |
| Haiku 4.5 | $0.00004 | $0.00555 |
Grade A, and why
hotspot-article 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 1mo 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 — 422 lines — stays where its author put it; the contents beside it link to each section on GitHub.
热点深度文章 Pipeline
这不是“一次提示词写长文”,而是一条有证据链和退出门槛的编辑流水线:
三轨机会池 → 需求优先评分 → 技术到业务桥接 → 多源研究 → 多视角提纲 → 实测 → 成稿 → 事实/技术/文风审校 → 质量门
对标腾讯云“内容精选”时,先读 docs/recipes/tencent-selected-benchmark.md,再按内容类型选质量档,不能用一个固定篇幅处理所有文章:
selected-analysis:重大事件或行业变化,6,000–12,000 有效字符;selected-framework:架构、组织或决策框架,5,500–11,000 有效字符;selected-hands-on:产品实测或项目实战,4,500–9,000 有效字符。
三类都控制在 5–9 个 H2。重量来自攻击链、分类框架、真实输入输出、失败修复、架构图和明确取舍,不来自拆出更多章节。selected-sharp 只用于技术短稿,不计入腾讯云精选批次;selected 保留为兼容旧稿。
机器分只衡量可审计的下限,不能把 8.6/10 直接解释为审美或洞察得分。最终发布仍需人工主编确认。
触发条件
- “追一下今天的 AI 热点,写成深度文章”
- “从客户项目和业务难题里找一个值得写的选题”
- “对标精选文章,做一篇 8–9 分的长文”
- “运行 hotspot-article”
如果用户没有给主题,同时扫描近期大事件、真实需求和常青决策三条线。默认周内容组合为:时事 30%、需求/解决方案 50%、常青框架 20%。选题先补齐业务场景,再按当前组合缺口选择;热度很高但说不清“谁在什么情况下要做什么决定”的内容,只进简报,不写精选文章。
把选题与评分写入 brief.md,不中断执行等待确认;如果涉及政治、医疗、金融、法律或未成年人等高风险主题,则停在选题报告,等待用户确认。
文件约定
ROOT=.prax/vault/hotspot-articles/<YYYY-MM-DD>/<topic-slug>
$ROOT/
├── brief.md
├── topic-candidates.json
├── opportunity-report.json
├── business-bridge.json
├── demand-brief.md
├── research-notes.md
├── source-index.json
├── outline.md
├── evidence.json
├── evidence/
├── draft.md
├── fact-check.json
├── fact-check.md
├── editorial-review.md
├── quality-report.json
└── publish-ready.md # 只有全部门槛通过后才创建
新文件使用 Write;修改已有文件使用 Edit。保留中间稿和失败报告,不覆盖研究证据。
Step 0:读取项目配置
可选配置 .prax/article.yaml:
audience: 中文开发者与技术决策者
topics: [AI Agent, LLM, RAG, 开源工具]
exclude: [纯融资传闻, 无原始出处的爆料]
lookback_hours: 48
candidate_limit: 20
tone: 克制、清晰、证据优先
target_mix:
event: 0.30
demand: 0.50
evergreen: 0.20
没有配置就使用以上默认值。当前日期和时区必须写入 brief.md。
Step 1:构建三轨机会池
每个候选标注一个 track:
A. 时事线 event
- 当天
.prax/vault/ai-news-hub/<DATE>/; - 先运行
python3 -m prax.article_integrations probe,把可用和缺失的适配器写入brief.md; TRENDRADAR_OUTPUT存在时,用python3 -m prax.article_integrations ingest-trendradar归一化其 Markdown、JSON 或 SQLite 输出;- AutoCLI 可用时,通过
collect-autocli抓登录态平台;扩展或登录态失败只跳过该源; - 官方博客、GitHub Trending、Hacker News、行业媒体;
- 正文优先使用 Crawl4AI;未安装时适配器会遵守
robots.txt并退回公开 HTTP 抓取。
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.
- 1mo ago First seen · 422 lines · 39 tokens per session scan A b074c3759947
hotspot-article is a skill published in the GitHub repository ChanningLua/prax-agent (272 stars, last pushed 1mo ago), licensed MIT. It adds 39 tokens to every session and 5,548 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-09-12.
hotspot-article compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| hotspot-articleChanningLua/prax-agent | Skill | 272 | 1mo ago | MIT | A | 39 | Claude Code |
| code-knowledge-graphvibeeval/vibecosystem | Skill | 539 | 2d ago | MIT | A | 24 | Claude Code |
| interview-transcript-editorur-grue/autopunk-media-skills | Skill | 43 | 2mo ago | MIT | A | 70 | Claude Code |
| deep-researchSECRET4422/mcp-deep-research-server | Skill | 1 | 23d ago | MIT | A | 64 | Claude Code, Codex |
| fact-checkerNickCrew/Claude-Cortex | Skill | 66 | 4mo ago | MIT | A | 71 | Claude Code, Codex |
| git-history-analyzerchainlesschain/chainlesschain | Skill | 11 | today | none | A | 28 | Claude Code |
| partitioning-strategy-advisorbookforge-ai/bookforge-skills | Skill | 42 | 6mo ago | CC-BY-SA-4.0 | A | 248 | Claude Code |