Sort

AgentEarth Tool Catalog

InformationRetrieval

356Tools
14tags
35Providers

Connect once to Google, Bing, Tavily, Exa, Firecrawl and more for web search, full-site crawling and structured extraction at lower cost.

  1. High-frequency retrieval and complex crawling are optimized for stability to reduce mid-task failures.
  2. Start search and crawling in one place for deep research, competitor monitoring, sentiment tracking and information collection.

Tool List

356 tools

ae_bid_get_top_suppliers

根据业务关键词筛选出Top中标供应商列表,统计中标次数、中标总金额、最近中标时间、中标省份/城市分布及合作采购单位。适用于分析特定业务领域的头部供应商、竞品分析或寻找合作伙伴。

ae_firecrawl_crawl

Crawl multiple related pages starting at a website URL, returning the crawl status and collected content. Parameters • Required url: starting URL string. All other top-level fields are optional. • Scope: includePaths/excludePaths (arrays of URL path patterns), limit (page limit), maxDiscoveryDepth (link-discovery depth), crawlEntireDomain (whole domain rather than starting path), allowSubdomains and allowExternalLinks. • Discovery and pacing: sitemap = skip/include/only; ignoreQueryParameters, deduplicateSimilarURLs, delay between requests, and maxConcurrency. prompt accepts natural-language crawl instructions. • scrapeOptions applies to each page. formats accepts markdown, html, rawHtml, screenshot, links, summary, changeTracking, branding, json, query and audio. Content controls include onlyMainContent, includeTags/excludeTags, removeBase64Images and redactPII. jsonOptions accepts prompt/schema; queryOptions requires prompt (up to 10,000 characters) and mode (directQuote/freeform; the schema declares freeform as its default). parsers supports pdf; pdfOptions.maxPages is an integer from 1 to 10,000. screenshotOptions accepts quality, fullPage and viewport; viewport requires width and height. Other options include waitFor, mobile, location.country/languages, proxy (basic/stealth/enhanced/auto), maxAge, storeInCache, zeroDataRetention, skipTlsVerification and profile (name is required when profile is provided; saveChanges is optional). lockdown is an exposed boolean whose meaning is not defined by the selected source. Do not assume unspecified defaults, time units or numeric ranges. Undeclared top-level or scrapeOptions fields are not accepted. Async behavior and results Submitting starts a background task and returns a task ID. After the task completes, use ae_get_result to retrieve the saved result; this does not start another crawl. The crawl result may contain id, url, status, data, next, total, completed, expiresAt, success, error and warning. data contains collected documents, with no fixed document structure promised here. next may provide a URL for another page of results, so do not assume all results fit in one response. A terminal result is not necessarily a successful crawl. Crawls can be slow and responses can exceed token limits. High page/depth limits can produce oversized output; low limits can miss pages. A /* starting-URL wildcard is not recommended. For one page use scraping; for tighter control consider map plus scrape. Safe mode is read-only: webhooks and interactive actions are unavailable.

ae_law_get_article

按一部法的名称和一条的条号,取这一条的原文、发文字号、时效和法宝链接。不是法规列表,也不是这一条历年怎么改。

ae_bilibili_fetch_video_subtitle

Retrieve subtitle information for a Bilibili video. Provide both required inputs as strings: a_id is the video ID and c_id is the CID of the target video content. For example: {"a_id":"114006081739452","c_id":"28400484458"}. The response contains status information and a data field for the subtitle-related result. The structure of data is not specified and it may be null, so the response does not guarantee a full transcript, subtitle track list, or download link. This tool retrieves subtitle information; it does not create subtitles or transcribe audio.

ae_jina_search_web

Enter your search query in the required `request_body.q` field. Jina AI Search returns matching pages in JSON, with URLs, titles, and extracted page text when available. The response can also include usage information. To read a specific webpage or PDF from a known URL, use Jina AI Reader instead.

ae_douyin_get_aweme_id

Extract a single Douyin video's identifier (aweme ID) from its link. Input: - url (required, string): The link to the video, for example https://www.douyin.com/video/7298145681699622182. Result: - On success, the response data contains the extracted video identifier. The exact data structure is not specified. - This tool extracts the identifier; it does not retrieve video details or download the video. Supply one video link per call.

ae_dataforseo_labs_google_keywords_for_site

Find keywords relevant to a specific domain using DataForSEO, including search volume, cost-per-click, and competition data.

ae_bid_aggregate_advanced

Advanced aggregation tool for multi-dimensional statistical analysis of bidding data. Supports grouping by time, geography, industry, brand, company type, or bid method, calculating metrics such as count, total amount, and average amount, and performing YoY/QoQ comparative analysis.

ae_bid_get_company_business_keywords

Query business keywords for a company by analyzing winning bid subject matters (sm_names) to understand its business scope, winning counts, total amounts, top clients, and regional presence.

ae_bilibili_fetch_video_play_info

Retrieves playback information for one video. **Input:** provide the required `url` as a video link. **Return:** the response data contains the video's playback information; the specific details are determined by the returned data.

ae_douyin_get_sec_user_id

## Purpose Extract a single Douyin user's sec_user_id from their profile link. ## Input - `url` (required, string): The user's Douyin profile URL, for example `https://www.douyin.com/user/MS4wLjABAAAANXSltcLCzDGmdNFI2Q_QixVTr67NiYzjKOIP5s03CAE`. - Provide one user profile link per call. Support for other link types is not documented. ## Result Returns the user's `sec_user_id`, an identifier for that user, rather than profile details or a list of users. The exact structure of the result data is not specified in the documentation.

ae_douyin_index_fetch_brand_suggest

Get autocomplete suggestions for a brand search. Input: keyword (required string), the brand-name keyword to search for, for example "泡泡玛特". Returns: matching brand suggestions, with information such as brand names and category IDs where available. The detailed fields of each suggestion are not specified. Use the results to choose a matching brand; they are not brand performance metrics or full brand profiles.

ae_sorftime_patent_search_from_image

Search for design patents related to a product image to support preliminary appearance-infringement risk screening before listing. Provide image and query_mode. Images may be HTTP/HTTPS URLs, Base64 data URIs, or raw Base64; use JPG/PNG under 4M, preferably with a clean background and the product occupying more than 80% of the image. Choose mode 1 (recommended) for intelligent association based on extracted features and outlines, including matching product photos to patent line drawings; choose mode 2 to match the image's pattern, shape, and color. Optionally narrow the search by patent issuing authority, LOC classification, legal status, product keywords (comma-separated), and applicant. Issuing authority, LOC classification, and legal status allow multiple selections. Returns up to 100 related patents. Results are for further review and do not establish infringement or guarantee freedom from infringement.

ae_deepgram_get_models_list

List Deepgram's latest public speech-to-text and text-to-speech models and their metadata. Set the optional include_outdated query parameter to true to include non-latest versions. Results contain stt and tts lists with model details such as names, identifiers, versions, languages, and architecture, plus capability or voice metadata where available. Custom models require the separate Get Project Models operation. This is a metadata lookup, not transcription or speech synthesis; it uses GET /v1/models and has no request_body or callback_method parameter.

ae_dataforseo_google_bulk_traffic_estimation

Estimate monthly organic, paid, featured snippet, and local pack traffic for up to 1,000 domains or webpages using DataForSEO Labs.