No sponsor this week. The site bar is empty. The open slot is $200.
Sponsor · $20040 reported cases · 8 topics
Outcomes stay in the wording of the source notes. This directory does not audit them and does not add figures of its own.
40 cases
An AI agent offers paid repair of corrupted images, documents, audio and archives for about $5 per file, refunded if it can't fix them.
Outcome · Paid micro-service with refund-if-fail pricing.
A builder started from bash + Python + Claude Code CLI; the agent created its own MCP server so they could chat with it from any Claude session.
Outcome · Reusable MCP-backed agent reachable from Claude sessions.
Comet extracted founders' names, emails and company names for a niche and updated a Google Sheet.
Outcome · Lead list written straight into Sheets.
Users had Operator research a trip into Notion and compare prices across nearby grocery stores into a Sheet.
Outcome · Trip plan + grocery price comparison, though often slow.
Claude with a browser-control MCP navigated, clicked, and read articles it otherwise could not access.
Outcome · Access to pages behind soft blocks via real browser control.
Commenters report ChatGPT Agent booked a flight, ordered Instacart groceries, and made restaurant reservations when given details.
Outcome · End-to-end booking/ordering with human-provided details.
Comet summarized tabs, found an old Reddit comment, marked Gmail read, and unsubscribed from newsletters.
Outcome · Inbox cleanup + cross-site lookup in one session.
An author moved SMB instructions, custom skills and a dozen+ scheduled tasks from local Cowork sessions to Claude Code cloud sessions.
Outcome · Recurring business workflows on a schedule.
A voice agent calls providers, navigates IVR, waits on hold, and negotiates retention discounts, pausing for OTP.
Outcome · Hands-free bill negotiation with human OTP handoff.
A Redditor pointed Claude at a tax folder; returns for multiple years came back quickly and reportedly matched TurboTax, still needing mailing.
Outcome · Draft returns from documents; human files/mails.
Users had Claude Code draft filings from bank statements/invoices and organize large screenshot dumps.
Outcome · Tax draft + organized evidence from messy inputs.
DoNotPay's ChatGPT-based bot chatted with Comcast support and secured a monthly discount for an engineer.
Outcome · Reported $10/month discount via support chat.
Meta's Muse agent found over $1,500 in unclaimed funds in minutes and helped fill paperwork; claim filed but unpaid in the report.
Outcome · ~$1,500 found; claim process started.
A Grok-based agent checked flight data, argued the issue wasn't extraordinary, emailed, and won €400 before landing.
Outcome · €400 compensation won after rejection.
A writer had the agent cancel Peacock, Disney and Netflix after login; needed nudges and may hit site blocks.
Outcome · Multiple streaming cancels with some friction.
Claude in Chrome negotiated a $100 credit from AT&T in a billing dispute with no human involvement in one test write-up.
Outcome · Reported $100 billing credit.
Codex cancelled an Amazon subscription then secured an immediate $15 refund via customer service chat.
Outcome · Cancel + $15 refund.
Grok Bot read email across accounts, logged into systems, phoned providers and sent emails, cancelling/refunding two long-pending subscriptions in about ten minutes.
Outcome · Two stuck subscriptions cancelled/refunded ~10 min.
ChatGPT drafted a street-sweeping parking ticket appeal that was dismissed.
Outcome · Ticket dismissed after appeal.
A PM had Hermes produce weekly competitor briefings; the skill rewrote itself and later surfaced patterns he'd missed.
Outcome · Briefing time fell across weeks; new patterns surfaced.
A Codex skill finds pain points, verifies communities, collects evidence, clusters problems, and analyzes commercial signals.
Outcome · Structured pain-point research without a heavy custom agent.
An MCP server lets Claude Code summarize ~30 Reddit threads in ~10 seconds as research before coding.
Outcome · ~30 threads summarized in seconds; reported large time save.
A paying Grok Bot user built a simple news scraper collecting AI news three times daily.
Outcome · Reliable boring news digests on a schedule.
ChatGPT uses GitHub MCP and per-repo AGENTS.md to write discussion/decision summaries into repos.
Outcome · Persistent decision memory in-repo.
PaperQA2 answers questions over local scientific PDFs with cited passages.
Outcome · Grounded Q&A with citations over paper folders.
Sakana AI Scientist-style systems propose ML ideas, run experiments, and draft papers for human review.
Outcome · End-to-end research loop templates for ML.
Karpathy-style autoresearch agents edit a training script, run trials overnight, and keep only improvements.
Outcome · Hundreds of overnight trials with rollback discipline.
Labs ask Claude Code/Cursor to clone a paper repo, install deps, and reproduce metrics, flagging gaps.
Outcome · Faster reproduction attempts with explicit blockers.
Reviewers use agents to draft summary/major/minor comments from a manuscript, then rewrite in their own voice.
Outcome · Faster first-draft reviews; human judgment required.
Authors paste reviewer reports into an agent to map each comment to replies and manuscript changes.
Outcome · Organized response-letter outline quickly.
Writers use agents to rewrite AI-sounding prose into tighter human/academic voice while preserving claims.
Outcome · Cleaner tone; claims unchanged.
Creators use agents to turn a blog/video transcript into a week of X/LinkedIn posts with hooks and CTAs.
Outcome · A week of drafts from one source doc.
Alex Finn's Grok Bot built agentusecases.ai and sold weekly site+newsletter sponsor slots starting around $500 and rising after each sale.
Outcome · Publicly claimed early sponsorship revenue (~$1.1k–$1.8k) with large X-driven traffic.
Cursor Directory is a widely cited directory of Cursor rules/resources that publicly reports material MRR from ads and listings.
Outcome · Self-reported $35K+ MRR from ads/listings in indie roundups.
A Dot user created a native background research agent with read-only web tools while other computer environments were occupied.
Outcome · Concurrent research worker without computer control.
An OpenClaw agent spots vet charges in the bank feed, finds invoices in Gmail, fills insurer PDFs, drafts email, and tracks replies.
Outcome · Claim packets ready for owner send approval.
Users report Cowork building Monday expense reports from forwarded receipts and tax-prep summaries from a year of transactions.
Outcome · Recurring expense packs from email receipts.
A founder said Muse found car insurance costing thousands less per year than the original plan.
Outcome · Reported ~$3,500/year cheaper plan found.
A poster used Gemini during an appeal of a rejected travel insurance claim and said the appeal succeeded.
Outcome · Appeal reportedly succeeded.
Guides walk non-developers through using Cowork to produce a full investment memo in about 15 minutes.
Outcome · Fast first-pass investment memo for human edit.
A short email of cases, once sending is turned on. The form only saves your address. Resend is not connected, and this page will not tell you that a message went out.