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_dataforseo_google_ads_search_volume

Get monthly search volume data for keywords from Google Ads via DataForSEO.

ae_bravesearch_llm_context

根据问题检索网页并返回相关正文片段及来源元信息,而不是生成最终答案。必填 q 用于提供问题或关键词;可选参数可调整参与筛选的结果数、纳入的网页数,以及每页和总 token 预算。结果包含 grounding.generic 中的网页 URL、标题和 snippets;sources 提供来源元信息(如请求启用且有返回)。总 token 限制为近似上限。

ae_github_repos_get

Get detailed information about a GitHub repository, including its owner, visibility, fork relationships (if it's a fork), and — when supported — its code of conduct.

ae_dataforseo_backlinks_referring_domains

Use DataForSEO to identify the unique domains linking to a specific website or page.

ae_law_get_case_list

按案号、标题、案由、当事人、裁判要点、罪名和筛选项,从北大法宝司法案例库检索案例列表。返回题录、法宝链接,以及库里已拆好的本院查明、本院认为、诉讼请求、争议焦点、裁判依据和裁判结果。不含整篇裁判文书正文。

ae_brightdata_search_dataset

Search Bright Data's existing LinkedIn datasets for people profiles, contact-enriched people profiles, or company information that match your criteria. Provide dataset_id and a nonempty filter. The supported IDs are gd_l1viktl72bvl7bjuj0 (people profiles), gd_me5ppxjr2ge6icjuh0 (contact-enriched people profiles), and gd_l1vikfnt1wgvvqz95w (company information). Before searching, call list_dataset_fields for the chosen dataset to learn valid field names and types. Build a filter from a single condition {name, operator, value}, or combine conditions with {operator:"and"|"or", filters:[...]}. Groups must be nonempty, with nesting limited to depth 3. Supported condition operators are =, !=, <, <=, >, >=, in, not_in, includes, not_includes, array_includes, not_array_includes, is_null, and is_not_null. Omit value for the two null checks; otherwise supply a value compatible with the field and operator. The optional size is 1–10 records, defaulting to 10. Use sort to select default ordering, random ordering, or custom field ordering with asc/desc. To retrieve another page, use default or custom ordering and supply the search_after cursor from the previous response. Returns matching Bright Data dataset records directly, without a snapshot step. Record fields depend on the chosen dataset and the service response; a fixed output schema is not supplied. This search operates on existing datasets; use run_scraper when you need fresh collection from specific URLs.

ae_firecrawl_map

Discover URLs on a website before choosing pages to scrape, or locate pages about a particular topic. This tool returns URL-discovery results, not page bodies. Parameters: - url (required URI string): website to map. - search (optional string): filter discovered URLs using a content or topic query. - limit (optional number): result limit; the confirmed schema specifies no default or numeric bounds. - sitemap (optional string): "include", "skip", or "only" to include, skip, or exclusively use the sitemap. - includeSubdomains (optional boolean): include URLs from subdomains. - ignoreQueryParameters (optional boolean): ignore query parameters when identifying and returning mapped URLs. No additional arguments are accepted. Defaults for optional inputs are not specified in the confirmed schema. Returns an object that may include links (discovered URLs), success (boolean or null), error (message or object), warning (string or null), id (string or null, map-run identifier for optional feedback), agent_hints (string array), and conversationId (string). None of these fields is required by the result schema, and the structure of individual links is not specified. conversationId is response metadata, not an accepted input here. Execution: a direct SimpleMapping request returns the result; this tool has no configured asynchronous submission, status polling, or separate result-retrieval workflow. The returned id is not a documented polling handle. Limitations: URL discovery does not extract page content or guarantee a complete inventory of a website. If you already know the target URL and need its contents, use a scraping tool. If a scrape gives empty or irrelevant content, mapping with search can help locate a better page. Example: ae_firecrawl_map({"url":"https://docs.example.com","search":"webhook events"}).

ae_bid_get_company_contacts

Query company contacts aggregated and extracted from bidding and tender announcements, providing phone numbers, contact names, appearance frequencies, and recent tender links.

ae_chanmama_video_understanding

Analyze the visual content and shot breakdown of one video. Supply exactly one video link or directly playable video URL in query.video_urls, and optionally specify analysis requirements in query.custom_prompt. Processing runs asynchronously.

ae_dumplingai_search_web

Search the web for organic results and snippets using DumplingAI.

ae_bid_get_timeline

Retrieves the full announcement timeline across all stages of a bidding project by providing either a bid notice ID with its type or a Zhiliao tender URL.

ae_dataforseo_backlinks_summary

Get a high-level overview of backlink data for a specific domain, subdomain, or webpage using DataForSEO.

ae_dataforseo_backlinks_backlinks

Retrieve backlinks and related metrics for a specific domain or webpage using DataForSEO.

ae_law_search_enforcement

检索北大法宝的行政执法文书(行政处罚决定书、行政处理、市场禁入等),返回一页可核对的文书题录和法宝链接,不返回文书全文。

ae_brightdata_search_scrapers

Find a Bright Data pre-built Web Scraper API scraper for a website or data type. Supply the required query string, such as "amazon reviews" or "linkedin.com"; optionally set limit to an integer from 1 to 50 (default 10). Results include matching scrapers' dataset_id, name, domain, and collection_methods. To check a selected scraper's input requirements, pass its dataset_id and one returned collection method to get_scraper_details.

ae_sorftime_patent_search_from_keyword

Search for patents using patent-related terms, a company or person name, or a publication or application number. A keyword query is required. Narrow the search by issuing authority, applicant, patent type, legal status, publication or application number, and patent classification; classification numbers are used together with a classification type. Results are paginated, with page 1 as the default and up to 30 records per page. Useful for researching competitor patent portfolios and gathering information for preliminary infringement-risk review; results do not establish a legal conclusion.