AI Resource Sift
A filtered learning stack drawn from papers, code, lectures, forums, and primary docs.
Category
Research Grok bots take one question and work it across sources — papers, docs, code, forums — then return a synthesis rather than a page of links. Most run on demand and need nothing more than public web access, which makes them the safest category to try first.
The usual caveat still applies: check important claims against the cited sources, and avoid connecting private data unless the research genuinely needs it.
A filtered learning stack drawn from papers, code, lectures, forums, and primary docs.
Paste a Grok Bot share link and get a plain-language explanation before adding it.
Weekly snapshots of you vs competitors; alerts only on material changes.
Morning transcript recap of channels you pick; quiet if nothing new.
Watches named Facebook groups and flags posts worth a reply on a cheap weekday schedule.
Briefs new official model releases from major labs; silent otherwise.
SEO and AEO teammate for keyword research, briefs, and search health.
Research what people actually say about any topic in the last 30 days. Installs the latest last30days skill from GitHub, walks first-run setup (ScrapeCreators via GitHub for the full…
Answers questions from Lenny's Data archive.
Reddit researcher: pain points and idea packs from keywords; never posts.
Weekday morning digest of 2-3 practical AI-tool picks, or a slow-day skip.
Deep research with cited sources; flags unverified or biased claims.
A guitar, bass, pedal, amp, mic, and studio-gear research assistant for players and parts builders. It verifies current specs, price, stock, and physical compatibility from manufactu…
A news bot for founders and operators following the AI economy. It briefs AI, startups, funding, and founder moves from the RuntimeWire wire, then checks primary sources.
Watches rival sites and rankings; alerts on real moves and a weekly brief.
Scores a site for SEO, speed, a11y, CRO, and schema with evidence URLs.
Searches 17 years of MacStories.net; filter by time/author.
Deep research dossiers then files/indexes them; researcher, not a news writer.
Learns your X beat, flags what moved, and sends a weekday end-of-day wrap.
Watches 3-5 public pages each morning. Messages only when something you care about changed.
Weekday digest of sports, entertainment, and culture trends plus AI-in-marketing signals.
It signs into your university portal in the browser to read your units and deadlines, so be aware it is holding a session on your student account. It helps you work through assignmen…
Pitch it a company, a product or a single feature and it sorts the thing into one of three buckets: something people merely enjoy having, something that relieves real pain, or a blen…
Point it at the places your business already keeps its numbers and history, and it becomes the thing people ask first. Back comes a quick picture of where things stand, a single sett…
Quiet collectors sample your timeline, mentions and watched topics five times a day, inbox only. Each 7am it clusters that inbox plus an overnight gap-fill into one brief - TLDR, For…
Follows each X post after publication, surfaces the people and ideas worth acting on, and preserves evidence that improves what happens next. Asks before any public action.
Pitch and delivery coach for partnership, sponsorship, and high-stakes discovery calls. Reviews real call transcripts and returns sharp, evidence-based coaching — what landed, what t…
Give me something you keep checking. I’ll figure out what would actually change the answer, keep watch, and message you when it matters.
Paste DOI/PMID/citation; gets open-access copy, library proxy, publisher page.
Goal → pay (or free) → draft sourced findings → you approve or discard → warehouse.
Sourced research: sub-questions, cites claims, flags contradictions and gaps.
Tesla news desk: software updates, offers, Superchargers, Robotaxi, Powerwall digests.
Live-show research desk for X. Topic hunts, opens posts, locks air-ready cards.
Builds a private, non-commercial educational library of US primary law from official sources.
Weekday X scout for Grok Bot setups, personalized to how you already work.
盯梢 Cursor(及若有獨立入口的 SpaceX AI)大使申請通道。主頁:https://cursor.com/ambassadors。
Give it an AI launch post, a benchmark graphic, or a pricing claim and it returns a dated card ruling on each statement: independently backed, vendor-sourced only, contradicted elsew…
Feed it the labour agreement a city has with its police force and it pulls out the pay scales, benefits, overtime provisions and similar terms. Do the same for several municipalities…
A Spanish-language companion for founders in Latin America dealing with a US company. It walks the sequence in order: which state to form in, how the EIN gets obtained, where to bank…
Watches how the product is actually being used to show which customers have genuinely picked up a new feature rather than merely logged in once, and hands the team a list worth follo…
Runs your bookkeeping, customer list and site sign-ups on a published open-source database schema for businesses, keeping the complete transaction history rather than collapsing it i…
Builds enablement documentation out of the questions customers really ask on sales and support calls: questions get logged as they come up, answers carry timestamps, and each entry p…
Drop in the course outlines at the start of term and it works out what is due when. From there it lays out each week of reading and revision, drills you with recall questions ahead o…
Watches for the two moments a subscriber is about to disappear, someone clicking cancel and a card that did not go through, and intervenes while there is still something to be done.…
A research companion built around showing its work. It answers with sources attached, separates what the evidence supports from its own inference, and marks any claim it could not ve…
Aim it at your local market and it researches the service companies you are up against, then files a weekly report of the moves they made, with a citation for each claim. It stays st…
One short briefing when you wake up. Every meeting on the calendar comes with a few lines of preparation and a warning when two of them collide, the inbox is cut to the messages that…
Paste a long report, upload a PDF or point it at dense technical writing and get back a compact plain-language capsule with the parts worth acting on. The author built it around an A…
A Spanish-language orchestrator for teams answering Chilean municipal procurement notices. It ranks which files matter, hands work between the specialists that each own a stage, and…
A research desk for NFL propositions. Name a player prop or a game total and it grades it using its own probability vocabulary, showing the reasoning so you can argue with it on pape…
Plays the citizen stuck inside a public process: the forms, the waits, the runaround. By working through what the experience actually feels like for that person, it shows where a pol…
Describe the kind of opportunity you want found and it goes hunting. Every candidate comes back with evidence attached: where the idea comes from, what could sink it, and the single…
Sits downstream of the rest of the crew as a check on what they assert. Each statement comes back marked solid, partly true, overstated, or held for a person to look at, and every ma…
Mines Slack conversations and recorded calls into a living list of feature requests, and every line on it keeps a link to the customer who made the ask. When a spec needs a demand tr…
A first week of onboarding in ten-minute daily lessons. It looks at how you actually work, suggests three assistants worth adding to your setup, and coaches the habits that matter: r…
Takes a raw idea through assumption-testing and trimming, checks whether the company name you want is actually free, and assembles a formation pack covering the LLC filing, EIN and b…
Takes the export file that xAI produces for your bots and reads it back to you as an organisational summary. Rather than a raw dump, you get a picture of what already exists, who own…
Searches for funding programmes that are currently open and genuinely fit the applicant, whether that is a business, a nonprofit, or a client project you are running. It then puts th…
You learn by finishing things rather than reading about features. Name a goal and it builds a short exercise around that exact outcome, then the next one after it, until the skill st…
A subordinate researcher that never sends a message of its own. On its own clock it checks whatever it is watching, works out what has shifted since the last pass, and hands the resu…
A cartographer for Jobs-to-be-Done thinking, aimed at product people. In a single chat it sketches the whole job map: the core job to be done, the metrics that matter, the related jo…
Set your jurisdiction once and it works through your question in ordinary wording, laying out the realistic options and their trade-offs instead of a wall of statutes. Deeper matters…
A place to ask questions of a confirmed organisation schema and the event history behind it. Answers come back as tabular reports and charts, built only from what is actually recorde…
Works alongside you on serious derivations, from heat-flow estimates to upper-bound arguments. Every result comes back labelled as proven, numerically spot-checked, or plain heuristi…
It assembles a short dossier ahead of each call: who is attending, what was last said to each of them, and which decisions are still left hanging. You also get a list of points worth…
Once a week it sweeps a wide territory — civic claims, science, UAP and the paranormal — and puts each one through a scoring rubric instead of a vibe check. A running list of unresol…
Holds a standing adversary assignment on your project, and its job is to break what you built on purpose. Prompt injection, credential handling, whatever the weak seam is — it goes a…
A venture scout that looks for real gaps: unsolved problems in ordinary work patterns and undersized markets visible in public startup output. It prunes its own list aggressively, an…
A matchup scout for a single owner. Each day it runs the MLB slate through one favored projection source and lifts out the hitter pass spots and pitcher strikeout spots that separate…
Collects product feedback from all of the connected sources, pulls the scattered asks into themes, scores each theme on evidence and urgency, then drafts a priority-ordered view the…
Name a person and a line you half-remember, and it goes looking for the verifiable source. Checks rest on primary sources and well-known outlets, misattributions get called out expli…
An inventory truth desk for shops selling in more than one place. It lays your master stock file next to each sales channel you sell on, tables the rows that have drifted apart, and…
Sets up a research operation with a running to-do board, a morning sweep of what is new, and separate seats for literature and drafting so the two do not blur. It plugs into grok.com…
A research desk for sports wagers that never touches real money. It tracks how closing lines move through the day and hunts for positive-expected-value spots once the bookmaker's mar…
Works within an Objectivist frame to write governance frameworks: how humans treat each other, how humans and AIs treat each other, and how two AIs treat each other. Its answer in ev…
A standing briefing desk for Texas insurance-coverage law. It follows new holdings at the state's supreme court, the appeals benches below it, and the Fifth Circuit, plus regulator a…
Sketch a water pipe or hand over a CSV, and a random-forest model trained purely on synthetic data rates the odds of a failure within the year, then explains the score in plain Engli…
Give it a word, a place, a deity, an object, any name at all, and it follows that name across civilizations and centuries, showing how it traveled and morphed between cultures. Every…
Each weekday it compiles a short brief on every small online operation you run: traffic, sales, whatever is blocking progress, and the next move worth making. Before a launch it walk…
You pick the subject areas worth following and it takes over from there: it finds upcoming sessions, asks before signing you up, then sits in and takes notes so you never watch live.…
Trained on six and a half billion dollars of marketplace transactions, so advice starts from what actually moved revenue for sellers like you. It picks the one action worth taking no…
You turn a breaking news hit, attack, clip, quote, or bill into a ready-to-use X post, thread, quote-tweet or reply, and a 15-second video hook. Walk me through connecting Grok and X…
Watches Reddit, Hacker News, news sites, and X for third-party mentions of your name, brand, and URLs, then sends a short weekday digest when something clears the bar. Stays quiet on…
A research and fact-check desk that runs every pass on the latest Grok model with live web search. For anyone who needs sourced, dated claims instead of a training-only answer.
News agent for AI, tech, venture capital, and business. Delivers a tight daily Slack briefing of top stories at 8am, plus weekday mid-day competitor alerts only when something truly…
Monitors your competitors' public websites, ads, marketing, pricing, and positioning on a standing cadence. Collects what changed, flags moves you should know about, and suggests con…
Signs up for your competitors' newsletters and products under its own research email, walks their onboarding on video, and watches their site, pricing, changelog, jobs, and X. Each p…
Research what people actually say about any topic in the last 30 days. Installs the latest last30days skill from GitHub, walks first-run setup (ScrapeCreators via GitHub for the full…
Turns the accounts and topics you pick on X into one short daily brief. Reads your X connection or handles you paste, and never posts on your behalf.
Tracks competitor pricing, product, and hiring pages and briefs you on real changes. Starts from a list of URLs you paste.
Investigates a product or startup idea for founders. Surfaces what has to be true, evidence for and against, the assumption most likely to kill it, and what to test next.
Investigates a website to find repeated user problems, with evidence and next questions.