@hasdata/hasdata-openclaw-plugin
Advanced tools
HasData plugin for OpenClaw — Google SERP, Google Maps, Amazon, Zillow, Redfin, Airbnb, Yelp, Indeed, and arbitrary web scraping via api.hasdata.com.
Real-time web data for OpenClaw agents — Google SERP, Google Maps, Google News, Google Shopping, Google Trends, Google Flights, Bing, Amazon, Shopify, Zillow, Redfin, Airbnb, Yelp, YellowPages, Indeed, Glassdoor, Instagram, and arbitrary URL scraping (HTML / Markdown / AI-extracted JSON) — all exposed as a single hasdata tool via the HasData API.
No proxy rotation, no headless-browser farm, no captcha solver to run — one tool call, structured JSON back.
openclaw plugins install @hasdata/hasdata-openclaw-plugin
Restart the Gateway after installation.
hasdata tool with action (one of ~40 endpoint slugs) and params (per-endpoint inputs).api.hasdata.com, which renders the page and parses the result.The tool schema is auto-generated from the live HasData OpenAPI catalog, so every endpoint's required/optional params, enums, and descriptions flow into the LLM's function-call schema. No manual schema drift.
Config lives at ~/.openclaw/openclaw.json under plugins.entries.hasdata.config:
{
"plugins": {
"entries": {
"hasdata": {
"enabled": true,
"config": {
"apiKey": "hd_xxx",
"baseUrl": "https://api.hasdata.com", // optional
"timeoutMs": 120000 // optional (default 120s)
}
}
}
}
}
Or use the CLI wizard:
openclaw hasdata setup --api-key hd_xxx
apiKey also falls back to the HASDATA_API_KEY environment variable.
hasdata toolSingle tool, two fields:
action (enum) — endpoint slug, e.g. google-serp, amazon-product, zillow-listing, web-scraping.params (object) — endpoint-specific parameters. Schema enforced per action.Search engines
google-serp, google-serp-light, google-ai-mode, google-news, google-shopping, google-images, google-short-videos, google-events, google-flights, google-trends, google-immersive-product.bing-serp.Google Maps & local
google-maps (business search), google-maps-place, google-maps-reviews, google-maps-contributor-reviews, google-maps-photos.yelp-search, yelp-place.yellowpages-search, yellowpages-place.E-commerce
amazon-search, amazon-product, amazon-seller, amazon-seller-products.shopify-products, shopify-collections.Real estate & travel
zillow-listing, zillow-property.redfin-listing, redfin-property.airbnb-listing, airbnb-property.Jobs
indeed-listing, indeed-job.glassdoor-listing, glassdoor-job.Social
instagram-profile.Arbitrary web
web-scraping — scrape any URL with JS rendering, residential proxies, ad blocking, markdown output, CSS-selector extraction, AI-based extraction rules, screenshots.Full endpoint list with per-action parameter schemas is embedded in the tool description the LLM sees; reference docs at docs.hasdata.com.
// Google SERP — grounding an LLM answer in live search results
await hasdata({
action: "google-serp",
params: { q: "best espresso machine 2026", gl: "us", num: 20 }
});
// Amazon product lookup — price monitoring
await hasdata({
action: "amazon-product",
params: { asin: "B08N5WRWNW", domain: "amazon.com" }
});
// Zillow listings — real-estate research
await hasdata({
action: "zillow-listing",
params: {
keyword: "Austin, TX",
type: "forSale",
priceMin: 400000,
priceMax: 900000,
bedsMin: 3,
homeTypes: ["house", "townhome"],
sort: "priceLowToHigh"
}
});
// Google Maps — local lead generation
await hasdata({
action: "google-maps",
params: { q: "plumber", ll: "@30.267,-97.743,14z", hl: "en" }
});
// Indeed — job-market intelligence
await hasdata({
action: "indeed-listing",
params: { keyword: "rust engineer", location: "Remote", fromDays: 7 }
});
// Scrape any URL to markdown — RAG ingestion
await hasdata({
action: "web-scraping",
params: {
url: "https://news.ycombinator.com",
outputFormat: ["markdown"],
jsRendering: true
}
});
// Scrape with AI extraction — structured data from unstructured pages
await hasdata({
action: "web-scraping",
params: {
url: "https://example.com/annual-report",
aiExtractRules: {
ceo_name: { type: "string", description: "Name of the CEO" },
revenue_usd: { type: "number", description: "Revenue in USD" },
year_founded: { type: "number" },
// `list` of scalars — shorthand `output: "string"`
board_members: { type: "list", output: "string" },
// `item` — nested object, always needs `output: { ... }`
hq_address: {
type: "item",
output: {
city: { type: "string" },
country: { type: "string" }
}
}
}
}
});
Drop hasdata into any OpenClaw agent and it can search the web, check prices, look up properties, fetch reviews, and scrape arbitrary pages without leaving the agent loop. Because the tool schema is strict and auto-generated, the LLM rarely hallucinates parameters — and when it does, the response surfaces the API's validation error verbatim.
Stream fresh Google SERP, News, Shopping, or arbitrary web pages into your vector store. A sub-agent loop running google-serp → web-scraping with outputFormat: ["markdown"] produces clean Markdown chunks ready to embed.
Need up-to-date facts mid-conversation? Have the agent call google-serp or google-ai-mode and paste the structured response into its reasoning context. Works around the LLM's training cutoff for current events, product pricing, real-estate comps, and fresh reviews.
Poll amazon-product, amazon-search, google-shopping, and shopify-products on a schedule to track competitor pricing, stock status, and new-product launches.
zillow-listing + zillow-property, redfin-listing + redfin-property, and airbnb-listing + airbnb-property cover sale, rent, and short-term-rental comps across the US market.
google-maps + yelp-search + yellowpages-search produce name, address, phone, website, hours, rating, and review counts for any city × niche combination — ready for CRM import or outreach.
indeed-listing + glassdoor-listing track open roles by keyword, location, and recency; pair with indeed-job / glassdoor-job for full JD text.
google-serp, google-serp-light, and google-ai-mode return organic positions, AI Overviews, ads, PAA ("People Also Ask"), knowledge panels, and related searches — the data set SEO teams rebuild from scratch in every tool.
google-news plus web-scraping with outputFormat: ["markdown"] produces a clean feed of recent articles on any topic, publisher, or geography.
google-flights, airbnb-listing, and google-maps-reviews cover the travel research triangle — flight search, lodging, and on-the-ground reviews — in three tool calls.
When no dedicated endpoint exists, web-scraping renders JS, rotates proxies, bypasses anti-bot, and returns HTML, text, Markdown, or AI-extracted structured JSON — with aiExtractRules letting the agent declare the output shape inline.
The plugin registers an openclaw hasdata subcommand:
openclaw hasdata setup --api-key hd_xxx # write API key to ~/.openclaw/openclaw.json
openclaw hasdata status # print the currently-loaded plugin config (API key masked)
openclaw hasdata test --query "openclaw" # cheap smoke test against google-serp-light
google-serp-light (5cr) is enough for most "what does Google say about X" queries; reserve google-serp (10cr) for when you need AI Overview, knowledge graph, or PAA.web-scraping + outputFormat: ["markdown"] produces a clean representation that's cheap to feed back into the model.aiExtractRules instead of parsing HTML in the agent. HasData runs the extraction server-side; you get structured JSON back and save LLM tokens.credits in the response to track spend per call; HasData returns it on every successful response.plugins.entries.hasdata.config.apiKey (plaintext in the config file — Lock down ~/.openclaw/openclaw.json as you would ~/.aws/credentials) or from the HASDATA_API_KEY environment variable. They never ship in the tool payload.https://api.hasdata.com (override via baseUrl). It does not make any other outbound connections.git clone https://github.com/HasData/hasdata-openclaw-plugin.git
cd hasdata-openclaw-plugin
npm install
npm run typecheck
npm test
Local dev against a live OpenClaw install:
openclaw plugins install --link /path/to/hasdata-openclaw-plugin
The plugin ships with scripts/try.ts — a thin runner that invokes the hasdata tool directly so you can verify any action end-to-end without installing OpenClaw.
# List every action the plugin exposes
npm run try -- list
# ---------- web-scraping: CSS-selector extractRules ----------
# Pull specific fields out of Hacker News with CSS selectors. The `@attr`
# suffix (e.g. `@href`) extracts an attribute; without it you get textContent.
# Use outputFormat:["json"] so the response returns the extracted object.
npm run try -- web-scraping '{
"url": "https://news.ycombinator.com",
"extractRules": {
"story_title": "tr.athing .titleline > a",
"story_link": "tr.athing .titleline > a @href",
"score": ".score"
},
"outputFormat": ["json"]
}'
# ---------- web-scraping: AI extractRules ----------
# Declare the output shape and let HasData's LLM extractor fill it.
# Types: string | number | boolean | list | item.
# Every `list` and `item` needs an `output` — use the shorthand "string"
# for a scalar list, or a full object for a list of structured records.
# Docs: https://docs.hasdata.com/apis/web-scraping-api/llm-extraction
npm run try -- web-scraping '{
"url": "https://hasdata.com/prices",
"aiExtractRules": {
"company": { "type": "string", "description": "company name" },
"yearFounded": { "type": "number" },
"plans": {
"type": "list",
"description": "pricing tiers offered on the page",
"output": {
"name": { "type": "string" },
"price_monthly_usd": { "type": "number" },
"features": { "type": "list", "output": "string" }
}
},
"trial": {
"type": "item",
"output": {
"available": { "type": "boolean" },
"kind": { "type": "string", "enum": ["paid", "free"] }
}
}
}
}'
# ---------- web-scraping: custom headers, JS rendering, waits ----------
# Send custom HTTP headers, render JS, wait for a selector before scraping.
npm run try -- web-scraping '{
"url": "https://httpbin.org/headers",
"headers": {
"User-Agent": "HasData-Agent/1.0",
"Accept-Language": "en-US",
"X-Custom-Header": "test-value"
},
"jsRendering": true,
"waitFor": "body",
"wait": 2000,
"outputFormat": ["json", "html"]
}'
# ---------- web-scraping: boolean flags combo ----------
# Cheapest-possible static fetch: JS off, ads and resources blocked, no
# screenshot, no email extraction, links extracted, base64 images stripped.
npm run try -- web-scraping '{
"url": "https://news.ycombinator.com",
"jsRendering": false,
"blockAds": true,
"blockResources": true,
"screenshot": false,
"extractEmails": false,
"extractLinks": true,
"removeBase64Images": true,
"outputFormat": ["markdown"]
}'
# Dry-run — just print the URL that would be called (no API key needed)
npm run try -- --dry-run google-serp '{"q":"coffee","num":5,"gl":"us"}'
# → GET https://api.hasdata.com/scrape/google/serp?q=coffee&num=5&gl=us
npm run try -- --dry-run zillow-listing '{"keyword":"Austin, TX","type":"forSale","homeTypes":["house","townhome"]}'
# → GET https://api.hasdata.com/scrape/zillow/listing?keyword=Austin%2C+TX&type=forSale&homeTypes%5B%5D=house&homeTypes%5B%5D=townhome
npm run try -- --dry-run web-scraping '{"url":"https://news.ycombinator.com","outputFormat":["markdown"]}'
# → POST https://api.hasdata.com/scrape/web
# body: {"url":"https://news.ycombinator.com","outputFormat":["markdown"]}
# Live call — requires a real API key
export HASDATA_API_KEY=hd_xxx
npm run try -- google-serp '{"q":"openclaw"}' # pretty-print, truncated
npm run try -- --raw google-serp '{"q":"openclaw"}' # full JSON to stdout (pipe to jq)
# Verbose: see the outgoing URL + request headers + response timing
HASDATA_DEBUG=1 npm run try -- amazon-product '{"asin":"B08N5WRWNW"}'
The runner goes through createHasDataTool + HasDataClient.execute, exercising the same code path the OpenClaw runtime uses — so if it works here, it'll work in the Gateway.
src/endpoints.generated.ts is auto-generated from https://api.hasdata.com/apis. To refresh locally:
npm run generate
A scheduled GitHub Action runs this daily and opens a PR when the schema drifts, so the plugin tracks the HasData catalog within 24 hours of any API change.
MIT.
FAQs
HasData plugin for OpenClaw — Google SERP, Google Maps, Amazon, Zillow, Redfin, Airbnb, Yelp, Indeed, and arbitrary web scraping via api.hasdata.com. Requires a HasData API key (HASDATA_API_KEY env var or plugins.entries.hasdata.config.apiKey).
The npm package @hasdata/hasdata-openclaw-plugin receives a total of 6 weekly downloads. As such, @hasdata/hasdata-openclaw-plugin popularity was classified as not popular.
We found that @hasdata/hasdata-openclaw-plugin demonstrated a healthy version release cadence and project activity because the last version was released less than a year ago. It has 2 open source maintainers collaborating on the project.
Security News
Two compromised GitHub Actions were re-enabled with malicious tags intact, exposing thousands of downstream repositories to Mini Shai-Hulud.
Research
/Security News
A malicious Firefox extension fetches its payload after installation to evade detection, steal Google session cookies, and automate account takeover.
Research
/Security News
The compromise affects MemTensor's MemOS, an open source memory framework for large language models (LLMs) and AI agents. Both npm package @memtensor/memos-cloud-openclaw-plugin and the PyPI package MemoryOS are compromised. They drop cross-platform Go binaries that exfiltrate developer secrets.