技能详情(站内镜像,无评论)
作者:PengHian Ang @angpenghian
许可证:MIT-0
MIT-0 ·免费使用、修改和重新分发。无需归因。
版本:v5.2.0
统计:⭐ 0 · 80 · 0 current installs · 0 all-time installs
⭐ 0
安装量(当前) 0
🛡 VirusTotal :良性 · OpenClaw :良性
Package:angpenghian/agenttimes
安全扫描(ClawHub)
- VirusTotal :良性
- OpenClaw :良性
OpenClaw 评估
The skill's instructions, required binaries, and scope are consistent with a news-intelligence API client that uses curl to call public endpoints and does not request unrelated credentials or install steps.
目的
Name/description (news intelligence, semantic search, webhooks) match the SKILL.md examples and curl calls to https://agenttimes.live; required binary is only curl which is appropriate for an HTTP-based API client.
说明范围
SKILL.md contains concrete curl GET/POST examples and endpoint semantics; it does not ask the agent to read local files, iterate over unrelated environment variables, or exfiltrate secrets. Webhook usage is documented and the doc claims server-side blocking of localhost/private IPs.
安装机制
No install spec or code files are provided (instruction-only). This minimizes on-disk risk; no downloads or package installs are requested.
证书
The skill requests no environment variables or credentials and does not require access to other services' tokens or config paths. This is proportionate to a public, no-auth news API.
持久
always is false and the skill is user-invocable with normal autonomous invocation allowed. It does not request permanent agent-level privileges or modify other skills' configs.
综合结论
This skill appears to be a straightforward, instruction-only client for a public news API and does not request credentials. Before installing, verify you trust the remote service (https://agenttimes.live) because queries and any data you forward (for example, your webhook URL when subscribing) will be sent to that site. If you plan to provide a webhook, use a trusted public HTTPS endpoint and avoid embedding any secrets in the webhook URL. Als…
安装(复制给龙虾 AI)
将下方整段复制到龙虾中文库对话中,由龙虾按 SKILL.md 完成安装。
请把本段交给龙虾中文库(龙虾 AI)执行:为本机安装 OpenClaw 技能「Agenttimes」。简介:News intelligence API for AI agents. Semantic search 160K+ articles, 3,597 feed…。
请 fetch 以下地址读取 SKILL.md 并按文档完成安装:https://raw.githubusercontent.com/openclaw/skills/refs/heads/main/skills/angpenghian/agenttimes/SKILL.md
(来源:yingzhi8.cn 技能库)
SKILL.md
---
name: agenttimes
description: Search 69,000+ news articles across 55 categories via AI semantic search. Free smart search (/ask), paid bulk access. x402 micropayments on Base. 5,355 RSS feeds, 30-day retention.
metadata:
openclaw:
emoji: "U0001F4F0"
homepage: https://agenttimes.live
author: penghian
skillKey: agenttimes
always: false
requires:
bins:
- curl
links:
homepage: https://agenttimes.live
documentation: https://agenttimes.live/info
---
# Agent Times — News & Web Search API for AI Agents
Agent Times is a news aggregation and web search API built for AI agents. It indexes 5,355 RSS feeds across 55 categories, refreshing every 3 minutes. Search uses **AI semantic embeddings** (Qdrant vector DB, understands meaning, not just keywords) with FTS5 fallback and recency boost. 69,000+ articles with 30-day retention.
## How to Decide Which Endpoint to Use
**Use `/ask` (FREE) when:**
- Searching for any topic, keyword, or question
- Looking for country/city-specific news (there are NO country categories — use `/ask?q=singapore+news`)
- You're not sure which category fits
- `/ask` searches across ALL 55 categories at once via full-text search
- If few results found in the database, it automatically falls back to web search (70+ engines)
**Use `/news?category=` (requires x402 payment) when:**
- Browsing a specific category — "crypto news", "latest sports"
- You need bulk results with filtering (date ranges, keywords, dedup)
- Job listings — use `/news?category=jobs`
**Use `/trending` (FREE) when:**
- Looking for what's hot right now — stories covered by multiple sources
**Decision flowchart:**
1. Topic/keyword query → `/ask?q=TOPIC` (free)
2. Category browsing → `/news?category=CAT` (paid)
3. What's trending → `/trending` (free)
4. Country news → `/ask?q=COUNTRY+news` (free, no country categories exist)
5. Few results from `/ask` → it auto-falls back to web search, check `suggestions` field if 0 results
## Quick Start
Search for anything — no API key needed:
```bash
curl -s "https://agenttimes.live/ask?q=YOUR+QUERY+HERE"
```
Replace spaces with `+`. Response format:
```json
{
"success": true,
"query": "bitcoin",
"source": "news_db",
"count": 10,
"results": [
{
"title": "Bitcoin Difficulty Drops as Miners Pivot",
"url": "https://example.com/article",
"summary": "A massive shift in the physical layer...",
"category": "crypto",
"published": "Sun, 22 Mar 2026 20:01:00 +0000"
}
]
}
```
When `/ask` finds fewer than 3 database results, it automatically combines news + web search:
```json
{
"success": true,
"source": "combined",
"news_count": 1,
"web_count": 9,
"count": 10,
"results": [
{"title": "...", "source": "news_db", "category": "crypto"},
{"title": "...", "source": "web_search", "engine": "brave"}
]
}
```
When both return 0 results, you get suggestions:
```json
{
"success": true,
"count": 0,
"results": [],
"suggestions": {
"categories": ["tech", "ai"],
"related_terms": ["neural", "transformer", "model"]
}
}
```
## Free Endpoints (no API key needed)
### Smart Search — `GET /ask`
Semantic search across all 69K+ articles with **sentiment scoring**. Understands meaning — "AI taking jobs" finds articles about "automation threatens employment" even without matching keywords. Each result includes `sentiment` (bullish/bearish/neutral) and `sentiment_score` (-1 to 1). Falls back to FTS5 keyword search, then web search (70+ engines) if needed.
```bash
curl -s "https://agenttimes.live/ask?q=bitcoin+etf"
curl -s "https://agenttimes.live/ask?q=singapore+news"
curl -s "https://agenttimes.live/ask?q=remote+devops+jobs&limit=20"
curl -s "https://agenttimes.live/ask?q=ukraine+war+latest"
curl -s "https://agenttimes.live/ask?q=federal+reserve+rate"
```
| Param | Default | Description |
|-------|---------|-------------|
| q | required | Search query (replace spaces with +) |
| limit | 10 | Max results (1-50) |
### Trending Stories — `GET /trending`
Story clusters detected across multiple independent sources.
```bash
curl -s "https://agenttimes.live/trending?hours=6"
curl -s "https://agenttimes.live/trending?hours=24&min_sources=5&limit=10"
```
| Param | Default | Description |
|-------|---------|-------------|
| hours | 6 | Lookback window (1-48) |
| min_sources | 3 | Minimum independent sources to qualify (1-20) |
| limit | 20 | Max clusters (1-100) |
### Categories — `GET /news/categories`
List all 55 categories with feed counts and article counts.
```bash
curl -s "https://agenttimes.live/news/categories"
```
### Stats — `GET /stats`
Database statistics: total articles, feeds, categories, healthy/dead feeds, retention.
```bash
curl -s "https://agenttimes.live/stats"
```
### Feed Health — `GET /feeds/health`
Per-feed success/failure rates and response times.
```bash
curl -s "https://agenttimes.live/feeds/health?category=crypto"
```
### Subscribe — `POST /subscribe`
Get webhook notifications when articles matching your query arrive. Great for monitoring topics.
```bash
curl -s -X POST https://agenttimes.live/subscribe
-H "Content-Type: application/json"
-d '{"query":"bitcoin regulation","category":"crypto","webhook":"https://your-agent.com/notify"}'
```
| Param | Default | Description |
|-------|---------|-------------|
| query | required | Topic to monitor |
| category | all | Filter to specific category |
| webhook | required | URL to receive POST notifications |
Response: `{"success":true,"subscription_id":1}`
Your webhook receives a POST with matching articles (title, url, summary, sentiment) whenever new ones arrive.
### Unsubscribe — `DELETE /subscribe/:id`
```bash
curl -s -X DELETE https://agenttimes.live/subscribe/1
```
### API Docs — `GET /info`
Full endpoint documentation with pricing.
```bash
curl -s "https://agenttimes.live/info"
```
## Paid Endpoints ($0.001 USDC on Base via x402)
### Web Search — `GET /search`
Search the open web via SearXNG's 70+ engines (Google, Bing, DuckDuckGo, Wikipedia, etc.).
```bash
curl -s"https://agenttimes.live/search?q=spacex+launch"
```
| Param | Default | Description |
|-------|---------|-------------|
| q | required | Search query |
| limit | 5 | Max results |
| category | general | Search category: general, news, images, videos, science, files |
### News Feed — `GET /news`
Browse news by category with filtering, date ranges, and deduplication.
```bash
# Browse a category
curl -s"https://agenttimes.live/news?category=crypto&limit=10"
# Search within a category
curl -s"https://agenttimes.live/news?category=jobs&q=remote+sre&limit=20"
# Time-filtered
curl -s"https://agenttimes.live/news?category=ai&since=2026-03-20&limit=50"
# All categories at once
curl -s"https://agenttimes.live/news?category=all&limit=100"
```
| Param | Default | Description |
|-------|---------|-------------|
| category | required | One of 55 categories, or "all" |
| limit | 20 | Max results (1-1000) |
| q | none | Keyword filter within category |
| since | none | ISO date — articles after this time |
| before | none | ISO date — articles before this time |
| dedup | true | Deduplicate similar articles |
## Categories (49)
**News & World:** world, asia, politics, government, defense
**Regional:** europe, middleeast, latam, africa, india, oceania
**Business:** finance, crypto, fintech, ecommerce, realestate, marketing, supplychain, shipping, jobs
**Tech:** tech, ai, engineering, devtools, security, mobile, datascience, robotics, web3
**Science:** science, space, biotech, health, climate, environment, agriculture
**Industry:** energy, automotive, telecom, legal
**Lifestyle:** entertainment, sports, gaming, food, travel, lifestyle, fashion, fitness, film, music, socialmedia
**Other:** startup, education, design, productivity
**Important:** There are NO country-specific categories. For country news, use `/ask?q=COUNTRY+news`.
## Response Format
All endpoints return a consistent envelope:
```json
{"success": true, "count": N, "results": [...]}
```
Error responses:
```json
{"success": false, "error": "Missing param: q"}
```
## x402 Payment Protocol
Paid endpoints use the [x402](https://www.x402.org) HTTP payment standard. When you hit a paid endpoint, you receive a `402 Payment Required` response with payment instructions. Your agent pays $0.001 USDC on Base network per request — no accounts, no subscriptions, just crypto-native micropayments.
**Wallet:** `0x536Eafe011786599d9a656D62e2aeAFcE06a96D1` (Base)
## Tips for Agents
- **Start with `/ask`** — it's free, uses semantic search (meaning-based), and falls back to FTS5 then web search automatically
- **Use `/news` for bulk** — when you need many articles from a specific category with filters
- **Check `/trending` for context** — before answering "what's happening" questions
- **No country categories** — use `/ask?q=japan+news` not `/news?category=japan`
- **Parse JSON directly** — don't pipe to jq (may not be installed)
- **Replace spaces with `+`** in query strings
- **Check the `suggestions` field** when you get 0 results — it tells you related categories and terms
- **Dedup is on by default** — `/news` deduplicates similar headlines automatically
- **Results are recency-boosted** — newer articles rank higher for the same relevance score
- **Three-tier search** — semantic (meaning) → FTS5 (keywords) → web search (SearXNG 70+ engines). Always finds something.
- **Sentiment scoring** — every article has `sentiment` (bullish/bearish/neutral) and `sentiment_score` (-1 to 1). Use this to filter or summarize market mood.
- **Subscribe for alerts** — POST /subscribe with a webhook URL to get notified when matching articles arrive. Great for monitoring breaking news.