Topic
Agents use cases: 442 examples
Source-linked agent workflows tagged Agents.
Muse ran organic app marketing over WhatsApp for a fresh account
Pedro Pérez onboarded Meta Muse on WhatsApp like a marketing hire for a quit-drinking app, had it research hooks, build slideshows, and schedule via Postiz; a brand-new account hit about 1.1M post views in days.
Grok Bot booked a hotel and flagged a duplicate stay before payment
X Freeze booked another hotel with Grok Bot; it required payment approval and noticed he already had a booking at the same hotel.
ChatGPT Agent fills a supermarket cart for next-morning pickup
From agentusecases.ai: Agent mode added groceries to a local supermarket cart while the user worked, ready for pickup the next morning—works best with meal-specific instructions.
Use Claude computer use to compile a native app and click-test every control
Claude Code's macOS computer-use preview can write Swift, compile/launch a menu-bar app, click through controls, and screenshot results in the same session—plus drive GUI-only tools and iOS Simulator when MCP/bash/browser tools aren't enough.
Build safer CDEs on AWS with template-defined agent workspaces
Workshop write-up: define persona templates (cloud-native, serverless, agentic, GenAI) as Terraform, GitOps them, and run parallel agents each in isolated AWS workspaces instead of conflicting on one laptop.
Run coding agents on a disposable VPS—one Linux user per agent
A Lobsters-linked write-up: isolate Claude Code/Codex-style harnesses on a cheap Hetzner VPS with named users (vikki/clara/…), tmux sessions, a shared bot GitHub account, and disposable machines so yolo-mode stays off your laptop.
ChatGPT workspace agent: daily sales meeting briefs from Calendar + SharePoint
OpenAI cookbook shows a schedulable ChatGPT workspace agent that finds external customer meetings, pulls SharePoint notes and news, writes a 2–3 page brief, saves it, and emails an exec summary.
OpenAI cookbook: Slack bot with Agents API, persistent sessions, sandboxes
Official Agents API cookbook: connect Slack threads to persistent agent sessions, workplace tools, and isolated sandboxes so workplace chat becomes a durable agent runtime.
n8n agent human fallback: when the bot doesn't know, Slack a human
Official n8n AI example: a chat agent answers with GPT-4; if unsure it calls a sub-workflow that asks for the user's email and posts the unanswered question to a Slack support channel.
n8n published agents: schedules, sub-agents, and Message-an-Agent in workflows
n8n docs describe first-class agents with tools, knowledge bases, hourly/daily schedules, and sub-agent handoffs—plus embedding agents inside fixed workflows via the Message an Agent node.
Kai: AI executive assistant across email, meetings, and calendar
Kai (by Morgen) triages inbox, drafts replies, captures meetings without a join-bot, and builds a weekday Daily Brief from calendar overlaps and tasks—send always requires your click; MCP available for other agents.
n8n first AI agent: chat trigger + weather/news tools in ~2 minutes
n8n’s starter template wires Chat Trigger, an AI Agent node, Gemini, weather + RSS tools, and short memory—so you prove tool-using agents before building the fancy multi-agent graph.
Gate refunds and DB writes behind deterministic checks, not prompt hopes
An Ask HN on controlling agents that take real actions converges on: LLM proposes → deterministic validator (rules, spend ceilings, replay checks) → human approve—especially for refunds, payments, and database writes where prompt 'never do X' fails.
Add a human trust gate before consequential multi-agent actions
After running nine agents that looped APIs overnight, a team argued for an agent control plane—checkpoint only when something consequential is about to execute, not a giant dashboard.
Statewright: state-machine guardrails so coding agents can't skip plan→test
Show HN: Statewright's Rust engine enforces plan/implement/test states for Claude Code (and other CLIs) via MCP hooks—wrong tools are blocked structurally, shrinking death spirals for both small and frontier models.
Turn repeatable agent research into an AgentRun workflow DSL
Show HN for AgentRun: split a research question into subquestions, fan out agents, screen evidence, then write the report—inspect each step without paying for a full agent loop every time.
Hotel that agents can book on the open web—no MCP required
Show HN: a hotel exposes live price/availability so AI agents can read rates and hand the user a booking link entirely over the open web, without MCPs or custom connectors.
When to build an agent: fraud, refunds, vendor review, insurance claims
OpenAI’s practical guide frames agents for complex judgment (fraud/refunds), brittle rule piles (vendor security reviews), and unstructured docs (home insurance claims)—not for simple deterministic checklists.
Anthropic: customer-support and coding agents with tool use and human oversight
Anthropic’s Building Effective Agents appendix highlights two proven patterns—support agents that pull data and take actions like refunds, and coding agents that iterate against tests—with clear success metrics and human review.
Interruptible Zapier+Claude: AI plans, control form reviews, then execute
Indie Hackers guide: Jotform intake → Zapier → Claude/ChatGPT triages priority and drafts a reply → a control form lets you edit → a second Zap finishes CRM/notify actions—only high-priority items pause for review.
Build a three-tier Claude marketing workforce that posts nothing without Slack approval
Indie Hackers founder Sean Prentice used Claude Code to build an orchestrator + marketing director + specialist agents (content/ads/research) for his SaaS; outputs land in Slack and nothing posts or spends without his explicit say (~$10/mo self-use).
Build an AI secretary around persistent context, not another chat box
An Indie Hackers post breaks down the architecture for an AI secretary that actually keeps context: session vs short-term task vs long-term preference layers so birthday/contact and calendar commitments survive across chats.
Cronloop: Markdown-defined agents that loop every 5 minutes to weekly
Product Hunt’s Cronloop lets you describe a recurring job in Markdown, pick Codex or Claude Code, connect tools, and run agents on a loop with durable memory that can self-improve each run.