Subnet teams' own channels, X posts, press, newsletters and events read from the chain — GitHub lives in its own tab
Desearch.ai | Subnet 22@desearch_aiOne agent task: "What has been published about open source AI agent frameworks since September 1?" One Desearch web search call with a start date and page text. Every result came back inside the date range, with the full page text. #searchAPI
Build a SERP API rank checker on GET /web. Store the best rank for each keyword as JSON and CSV, then diff a later run.
SN22 Desearch news pagesource ↗
Build a tweet search web app for public X posts. Map filters to GET /twitter, check the form, and download the same rows as CSV.
SN22 Desearch news pagesource ↗
Desearch.ai | Subnet 22@desearch_aiGive Desearch crawl a URL it can't fetch and it now tells you plainly that the URL is invalid. Less guessing when your agent hits a bad link. #searchAPI
Desearch.ai | Subnet 22@desearch_aiFull walkthrough with curl and a Python reply digest:
Desearch.ai | Subnet 22@desearch_aiAdd context with GET /twitter/post for the original post and GET /twitter/post/retweeters for who reposted it.
Desearch.ai | Subnet 22@desearch_aiPlan for the limit: replies come back as a sample, not the full thread. There is no cursor and no reply total, so don't treat the result as every reply.
Desearch.ai | Subnet 22@desearch_aicount sets how many replies to ask for, 1 to 100 (default 10). Add query to filter, for example a keyword or lang:en. No match means an empty array, not an error.
Desearch.ai | Subnet 22@desearch_aiGET /twitter/replies/post?post_id=... The post ID is the number after /status/ in the post URL. You get back an array of reply posts with text, user, url, and engagement fields.
Desearch.ai | Subnet 22@desearch_aiThe replies under a post often say more than the post itself. Here's how to pull replies to any public X post as structured JSON with one call 🧵
Desearch.ai | Subnet 22@desearch_aiBuilder tip: one Desearch web search call can ask for page text and a publish-date window together. Fresh pages, full text, one request. #searchAPI
Desearch.ai | Subnet 22@desearch_aiFull walkthrough with curl and a Python tool that returns page text next to its source URL:
Desearch.ai | Subnet 22@desearch_aiEvery billable response carries X-Desearch-Cost-Usd and X-Desearch-Usage-Count headers, so your agent can log what each search and each page cost.
Desearch.ai | Subnet 22@desearch_aiStep 2: GET /web/extract?url=... Returns the page content as plain text by default. Set format=html if you want the markup instead.
Desearch.ai | Subnet 22@desearch_aiStep 1: GET /web?query=... You get a data array of results with title, link, and snippet. Page through with start (skip count, up to 100).
Desearch.ai | Subnet 22@desearch_aiMost web agents run the same loop: find pages, then read them. With Desearch that is two GET calls. Here's the whole pattern 🧵
Pull replies to any public X post as structured JSON with GET /twitter/replies/post: find the post ID, set count, filter with query, and add post context.
SN22 Desearch news pagesource ↗
A two-call web search API tutorial for AI agents: find pages with GET /web, then pull clean page text with GET /web/extract. Real parameters, curl and Python examples.
SN22 Desearch news pagesource ↗
Desearch.ai | Subnet 22@desearch_aiNeed results from a specific window? Set a start and end date on Desearch web search and keep results inside that publish range. #searchAPI
We parsed 6,480 robots.txt files from the Tranco top 10,000. 18.6% block AI training crawlers site-wide, 2.4% block the main search crawler. Data, method, CSV.
SN22 Desearch news pagesource ↗
Desearch.ai | Subnet 22@desearch_aiFull guide with curl and a Python script that runs the whole flow:
Desearch.ai | Subnet 22@desearch_aiBonus: count the hashtags across results and you see which neighboring tags your audience uses. It's an easy way to widen a monitor.
Desearch.ai | Subnet 22@desearch_aiStep 3: comments Take the shortcode of the most discussed post and call GET /desearch/instagram/comments/{media_id} You get text, author, likes, replies.
Desearch.ai | Subnet 22@desearch_aiStep 2: hashtag media GET /desearch/instagram/hashtag/{tag} (no #) Each post comes back with caption, createTime, stats, and every hashtags entry in the caption.
Desearch.ai | Subnet 22@desearch_aiStep 1: keyword search POST /desearch/instagram/search Returns public profiles with username, plus followers and mediaCount when available. Good for "who's active on this topic?"
Desearch.ai | Subnet 22@desearch_aiWant an agent or dashboard to follow an Instagram hashtag? You need three calls: find the accounts, pull the posts, read the comments. All public data, all JSON 🧵
Desearch.ai | Subnet 22@desearch_aiDoes your agent need links to read, or a cited answer it can use right away? With the Desearch AI search API, one request can return either. The Desearch blog walks through both.
Validator hotkey 5CoZxgtfhcJKX2HmkwnsN18KbaT9aih9eF3b6qVPTgAUbifj announced new child keys on SN1, SN2, SN3, SN4, SN5, SN6, SN7, SN8, SN9, SN10, SN11, SN12 and 113 more, applying at block 9,281,605 (~24h). Stake delegated to it there routes through its children after that.
Between blocks 9,274,365 and 9,274,433
Search public Instagram profiles by keyword, pull hashtag media, and read the comments on one post. Public data as JSON, with REST examples in curl and Python.
SN22 Desearch news pagesource ↗
Desearch.ai | Subnet 22@desearch_aiDesearch search now returns the page text for ordinary queries. Your agent reads the page, not just a snippet. #searchAPI
