Log new YouTube uploads in a Google Doc each morning
A Make workflow checks a channel’s latest upload at 9 AM, summarizes its title and description with Hugging Face, and saves the summary in Google Docs.
Agent Uses · 1,321 real examples
Literature reviews, market scans, fact-finding and learning new topics fast. Each entry shows what the person asked for, how it went, and a prompt to try it yourself.
The day's AI news plus 3 agent uses you can try, with ready-to-paste prompts. Free; unsubscribe anytime.
A Make workflow checks a channel’s latest upload at 9 AM, summarizes its title and description with Hugging Face, and saves the summary in Google Docs.
An n8n flow collects Reddit threads, lets Gemini pull out themes, and posts the findings into Notion.
A user left Gemini to process 170k LLM logs overnight with the llm CLI and checked the finished analysis the next day.
A product manager had Hermes generate a recurring competitor briefing that got quicker and uncovered patterns he had missed.
Manus, linked to Heurist Mesh, researched US stablecoin adoption in 2026 and produced a full report in about two minutes.
Manus used current web research to build a project management pricing comparison and then cleaned up a long Python file without changing how it worked.
Manus can check Reddit, X, and Hacker News each morning, sort the findings, and email or save a short digest.
A data analyst had Claude Code assemble a five-part analysis system from a detailed specification.
An agent sampled jobs on five company boards and measured how long each posting had been open.
They used Perplexity Comet to collect X material, condense YouTube transcripts, and turn both into presentation-ready notes.
A student gave one assistant a name and used it for a week as a standing helper for research and everyday tasks.
A Grok bot monitored X and AI coverage, checked figures against sources, and delivered five polished research briefs each day.
A developer had five Grok Bots read, test, and verify new papers overnight, then assemble a brief for morning review.
A writer used a Claude swarm to gather thousands of bot names and compare patterns by gender, industry, and launch year.
They used Grok Bot agents to split research, analysis, and production into separate steps instead of relying on one chat thread.
When an indexer failed on an uploaded book, Claude used shell commands to identify the file type and search the text directly.
A non-engineer used Claude Code to pull equity research from saved filings and notes, then checked the numbers with a second pass.
A commenter had Claude Code summarize each file’s unusual methods so they could keep track of a project they were steadily revising.
They used scheduled agent runs to scan watchlists for fresh company news, events, and insider activity.
They used Codex with the Logseq CLI docs to query a note graph and turn months of journals and tagged pages into summaries.
A commenter used ChatGPT Agent as a visual web browser to check many stores for one item and see which ones would ship to their country.
A commenter uses a custom browser agent for a narrow set of research and data-entry jobs, including finding scientific papers and downloading bank CSVs.
A Claude Code agent with paper search found useful ideas that a plain run missed while tuning a small language model.
A SciSpace agent can draft PICO questions, search papers, screen them, extract data, and assemble a review with a PRISMA diagram.