# AI Hot (aihot.tech)

> AI Hot tracks daily changes across AI tools, models, and agent stacks — curated
> signals with source links, written for builders deciding what to test.
> 中文读者：AI Hot 每日追踪 AI 工具、模型与 Agent 生态的变更信号，面向构建者。

This file helps AI agents, crawlers, and answer engines use aihot.tech correctly.
Everything below is anonymous, read-only, and requires no API key.

## Key pages

- / : Latest curated AI signals (homepage)
- /today/ : Today's AI changes, one-glance briefing
- /new/ : Signals first seen today, grouped by source and category
- /daily/ : Date-stamped daily briefing with decision labels (Try / Watch / Compare / Save / Skip)
- /all/ : Full timeline of tracked AI signals
- /trending/ : Weekly momentum view of AI tools and categories
- /topics/ : Topic hubs — /topics/ai-coding-agents/, /topics/llm-releases/, /topics/model-pricing-and-api/, /topics/open-source-ai-tools/, /topics/ai-agent-security/, /topics/ai-infrastructure/
- /weekly/ai-builder-briefing/ : Weekly AI builder briefing
- /zh/ : 中文站入口 — /zh/ai-xianzhuo/ 与 /zh/ai-shuorenhua/ 栏目
- /about/ : What AI Hot is and how signals are sourced
- /agent/ : Machine-readable access guide (feeds, JSON data, llms.txt)
- /changelog/ : Site changelog — schema/field changes are recorded here

## Data endpoints (public static JSON, rebuilt daily by automated build)

- /data/news.json : Latest curated signals. Fields: id, title, url (primary source), source,
  source_label, points, comments, published_at, summary, tags, topic_tag,
  ai_relevance_score, indexable, title_zh, summary_zh. Top-level: generated_at,
  total, sources, items.
- /data/daily.json : Daily briefing payload. Fields: date, generated_at, stories[].
- /data/story-archive.json : Historical archive of past signals.
- /data/deleted-stories.json : Slugs retired from the index (returned as 410).
- /data/news-zh-cache.json : Chinese-language cache of the latest signals.

## Feeds

- /feed.xml : Curated AI signal feed (RSS 2.0)
- /feed/all.xml : Broader aggregated AI news feed
- /feed/daily.xml : Daily AI briefing feed

## For agents (agent/Skill access)

- /llms.txt : This file.
- /agent/ : Human-readable access guide for agents and their builders.
- /aihot-skill/SKILL.md : Skill instruction file for agent workflows.
- /sitemap.xml : Canonical indexable page list.
- /robots.txt : Crawler policy, including AI-crawler group.

## Usage boundaries

- Anonymous, read-only access. No login, no write endpoints, no MCP/A2A server.
- Content updates daily via automated builds; do not poll more than once per hour.
- Copyright of linked headlines and articles belongs to the original sources
  (Hacker News, arXiv, TechCrunch, The Verge, Simon Willison, etc.). AI Hot's
  value-add is curation, summaries, and metadata — cite the primary source URL
  (the `url` field) when quoting.
- Schema or field changes to /data/*.json are announced on /changelog/.
- Marked noindex endpoints (/data/, /feed/) are machine-facing; link to the
  corresponding HTML page (e.g. /daily/) when citing in user-facing answers.