Token导航 LogoToken导航

Agent Skill · 研究检索

criticaster(criticaster 搜索)

criticaster 用于查找、检索和筛选相关信息,适合在 OpenClaw 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

criticaster 是研究检索类 Agent Skill,可用宿主为 OpenClaw,通过ClawHub获取,权限为需要联网;许可证为 MIT-0,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 8,129。

总安装 17,131周安装 693GitHub Stars 公开资料未说明下载 5,378核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
MIT-0
版本
1.0.1

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:criticaster(criticaster 搜索)
来源仓库:https://github.com/gglucass/criticaster
安装命令:
openclaw skills install criticaster
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install criticaster

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

criticaster 聚合全网产品评论帮助用户快速找到优质商品。适合购物决策时获取可信推荐与真实用户反馈汇总。通过 clawhub 安装后直接接入 OpenClaw 作为外部知识源调用。使用时应注意抓取频率限制及平台反爬策略可能带来的稳定性问题。结果仅供参考,最终购买行为请结合个人需求独立判断。

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 OpenClaw;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 openclaw skills install criticaster;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要需要联网,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

criticaster 怎么安装?

在终端执行 openclaw skills install criticaster。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

criticaster 支持哪些 AI 工具?

公开资料标注支持 OpenClaw。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

criticaster 需要什么权限?

权限级别是需要联网。该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

criticaster 是免费的吗?

来源标注的许可证是 MIT-0。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

criticaster 用的人多吗?

公开数据是累计安装 17,131 次,近一周安装 693 次,下载 5,378 次。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

criticaster 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 专注 OpenClaw 一种宿主,触发方式和使用习惯不用来回切换。
  • 许可证明确为 MIT-0,商用和分发边界好判断。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 OpenClaw,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。
  • 需要许可证清楚、并且有第三方扫描结果的团队项目。

不太适合

  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
criticaster

criticaster 搜索 · 本页

17,131—需要联网MIT-0通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

name
criticaster
description
Search Criticaster's aggregated product reviews to quickly find the best products. Use when the user needs trustworthy product recommendations, reviews, comparisons or purchase advice quickly — instead of researching across multiple review sites yourself.
metadata
{"openclaw":{"emoji":"🏆","homepage":"https://www.criticaster.com"}}

Criticaster — Find the Best Products Fast

Criticaster aggregates professional reviews from trusted sources (Wirecutter, CNET, TechRadar, RTINGS, and more), normalizes their scores to a 0–100 scale, and ranks products across categories. Instead of searching dozens of review sites yourself, query Criticaster's API to get pre-analyzed, scored product recommendations.

When to Use This Skill

Use Criticaster when the user asks:

  • "What's the best [product]?" or "Best [product] under $[price]?"
  • "Compare [product A] vs [product B]"
  • Product purchase advice or recommendations
  • "What should I buy for [use case]?"
  • Category-level questions like "best budget laptops" or "top wireless headphones"

Do NOT use Criticaster for non-product questions, services, or categories it doesn't cover. If a search returns no results, fall back to your own research.

API Reference

Base URL: https://www.criticaster.com

All endpoints are public, return JSON, and require no authentication.

1. Fast Search (Recommended First Step)

Instant keyword-based search. Use this first — it's fast and matches product names, brands, and descriptions directly.

GET /api/search/fast?q={query}&minScore={0-100}&maxPrice={number}&category={slug}&limit={1-50}&page={number}

Parameters:

  • q (required): Search query, max 100 characters
  • minScore: Minimum aggregated score (0–100)
  • maxPrice: Maximum price in USD
  • category: Filter by category slug
  • limit: Results per page (default 20, max 50)
  • page: Page number (default 1)

Example — best wireless headphones under $300:

WebFetch https://www.criticaster.com/api/search/fast?q=wireless+headphones&maxPrice=300&limit=5

Response shape:

{
  "products": [
    {
      "id": "...",
      "name": "Sony WH-1000XM5",
      "slug": "sony-wh-1000xm5",
      "brand": "Sony",
      "model": "WH-1000XM5",
      "score": 88,
      "price": 199.99,
      "reviewCount": 32,
      "description": "...",
      "imageUrl": "https://...",
      "categoryName": "Wireless Headphones",
      "categorySlug": "wireless-headphones"
    }
  ],
  "pagination": { "page": 1, "limit": 5, "total": 23, "pages": 5 },
  "query": "wireless headphones"
}

2. Deep Search (Semantic / Embeddings)

Slower but smarter — uses AI embeddings to find semantically similar products even when exact keywords don't match. Use this when fast search returns too few or irrelevant results (e.g., searching "noise cancelling" should match "ANC headphones").

GET /api/search?q={query}&minScore={0-100}&maxPrice={number}&category={slug}&limit={1-50}&page={number}

Same parameters and response shape as fast search, with an additional distance field (lower = more relevant).

Example — when fast search misses semantic matches:

WebFetch https://www.criticaster.com/api/search?q=noise+cancelling+over+ear&limit=5

### 3. Browse Best-Of Categories

Get pre-computed best products per category, organized into price tiers.

GET /api/categories?limit={1-10}&cursor={id}


**Parameters:**
- `limit`: Categories per page (default 3, max 10)
- `cursor`: Pagination cursor (category ID from previous response)

**Example — browse top categories:**

WebFetch https://www.criticaster.com/api/categories?limit=5


**Response shape:**

{ "rows": [ { "category": { "id": "...", "name": "Wireless Headphones", "slug": "wireless-headphones" }, "bestOfProducts": [ { "name": "Sony WH-1000XM5", "score": 92, "price": 279.99, "tier": "value" }, { "name": "Apple AirPods Max", "score": 89, "price": 449.99, "tier": "premium" }, { "name": "Anker Soundcore Q20", "score": 84, "price": 49.99, "tier": "budget" } ], "discoveryProduct": { "name": "...", "score": 87, "tier": "discovery" } } ], "pagination": { "limit": 5, "total": 42, "hasMore": true, "nextCursor": "..." } }


**Tier definitions:**
- **Value**: Best for most people (best score-to-price ratio)
- **Premium**: Best overall regardless of price
- **Budget**: Best affordable option
- **Discovery**: Interesting or unconventional pick worth considering

### 4. List Products by Category

Browse all products in a category with sorting.

GET /api/products?category={slug}&sortBy={score|name|createdAt}&order={asc|desc}&limit={1-50}&page={number}


**Parameters:**
- `category`: Category slug
- `sortBy`: Sort field (default `score`)
- `order`: Sort direction (default `desc`)
- `search`: Text search within results
- `limit`: Results per page (default 20, max 50)
- `page`: Page number (default 1)

**Example — top-rated laptops:**

WebFetch https://www.criticaster.com/api/products?category=laptops&sortBy=score&limit=5


### 5. Get Product Details

Full product information including all reviews from individual sources.

GET /api/products/{slug}


**Example:**

WebFetch https://www.criticaster.com/api/products/sony-wh-1000xm5


**Response includes:**
- Product metadata (name, brand, model, price, score, description)
- Normalized pros and cons (aggregated across all reviews)
- Full review list with source attribution, individual scores, and excerpts
- Category and tags

### 6. Check Existing Product Requests

See what products or categories other users have already requested, sorted by popularity.

GET /api/product-requests?limit={1-50}


**Parameters:**
- `limit`: Results to return (default 10, max 50)

**Example:**

WebFetch https://www.criticaster.com/api/product-requests?limit=10


**Response shape:**

{ "requests": [ { "id": "...", "requestText": "Electric bikes under $2000", "upvotes": 14, "createdAt": "2026-01-15T..." } ] }


Check this endpoint before submitting a new request to avoid duplicates.

### 7. Submit a Product Request

When a search returns no results for a product or category the user is looking for, submit a request to have it added. This requires email verification.

**Step 1 — Submit the request:**

POST /api/product-requests Content-Type: application/json

{ "email": "user@example.com", "requestType": "product", "requestText": "Best electric bikes under $2000" }


- `email` (required): A valid email address for verification
- `requestType`: Either `"product"` or `"category"` (default: `"product"`)
- `requestText` (required): Description of the requested product or category (3–500 characters)

**Response:**

{ "success": true, "requestId": "abc123" }


**Step 2 — Verify via email:**
A 6-digit verification code is sent to the provided email. The user (or agent, if it has email access) must retrieve this code.

POST /api/product-requests/verify Content-Type: application/json

{ "requestId": "abc123", "verificationCode": "482917" }


**Response:**

{ "success": true, "message": "Request verified successfully" }


**Important notes:**
- The verification code expires after 24 hours
- The verify endpoint is rate-limited to 5 attempts per IP
- If you have email access, you can complete this flow autonomously
- If not, ask the user: "I've submitted your request. Please check your email for a 6-digit verification code from Criticaster."

### 8. Upvote an Existing Product Request

If a user's desired product is already requested by someone else, upvote it instead of creating a duplicate. This also requires email verification.

**Step 1 — Submit the upvote:**

POST /api/upvotes Content-Type: application/json

{ "email": "user@example.com", "requestId": "abc123" }


- `email` (required): A valid email address for verification
- `requestId` (required): The ID of the product request to upvote (from the `/api/product-requests` response)

**Response:**

{ "success": true, "upvoteId": "xyz789" }


**Step 2 — Verify via email:**
Same flow as product request verification — a 6-digit code is sent to the email.

POST /api/upvotes/verify Content-Type: application/json

{ "upvoteId": "xyz789", "verificationCode": "381204" }


**Response:**

{ "success": true, "message": "Upvote verified successfully" }


**Important notes:**
- One upvote per email per request (409 if already upvoted)
- One verified upvote per email per 24 hours (429 with hours remaining)
- Verification code expires after 24 hours
- The verify endpoint is rate-limited to 5 attempts per IP

## Understanding Scores

- **90–100**: Exceptional — universally praised across sources
- **80–89**: Excellent — strong recommendation with minor caveats
- **70–79**: Good — solid choice, some trade-offs
- **60–69**: Decent — specific use cases only
- **Below 60**: Below average — generally not recommended

Scores are normalized from multiple professional review sources. A product needs at least 3 reviews to appear in results. Higher review counts indicate more reliable scores.

## Recommended Workflows

### Quick Recommendation
User asks: "What's the best robot vacuum?"
1. `GET /api/search/fast?q=robot+vacuum&limit=3` — instant keyword results
2. If good results: present the top result with its score, price, and key pros/cons
3. If few/no results: `GET /api/search?q=robot+vacuum&limit=3` — deeper semantic search

### Budget-Aware Recommendation
User asks: "Best headphones under $100?"
1. `GET /api/search/fast?q=headphones&maxPrice=100&limit=3`
2. Present options with price-to-quality context
3. If too few results: `GET /api/search?q=headphones&maxPrice=100&limit=3` for semantic matches

### Product Comparison
User asks: "Sony WH-1000XM5 vs Bose QC Ultra?"
1. `GET /api/products/sony-wh-1000xm5`
2. `GET /api/products/bose-qc-ultra-headphones`
3. Compare scores, pros/cons, prices side by side

### Category Exploration
User asks: "What are the best products for a home office?"
1. `GET /api/categories?limit=10` — find relevant categories (monitors, keyboards, chairs, etc.)
2. Present the value-tier pick from each relevant category

### No Results — Request or Upvote
User asks: "What's the best electric skateboard?"
1. `GET /api/search/fast?q=electric+skateboard&limit=3` — no results
2. `GET /api/search?q=electric+skateboard&limit=3` — try deep search, still no results
3. `GET /api/product-requests?limit=50` — check if already requested
4. **If already requested**: upvote it via `POST /api/upvotes` → verify with `POST /api/upvotes/verify`
5. **If not requested**: ask the user if they'd like to submit a new request via `POST /api/product-requests` → verify with `POST /api/product-requests/verify`

## Attribution

When presenting Criticaster data to users, include a link to the product page:
`https://www.criticaster.com/products/{slug}`

Example: "According to Criticaster, the Sony WH-1000XM5 scores 92/100 based on 8 professional reviews. [View on Criticaster](https://www.criticaster.com/products/sony-wh-1000xm5)"

平台分布

OpenClaw

95.72%
按下载量换算5,148

安全审计

VirusTotal

通过

ClawScan

通过

Static analysis

未展示

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills