Superbash
Search

Practical AI tutorials, organized by topic.

Browse 43 step-by-step tutorials for learning Grok Bot, running Hermes Agent, choosing models, and building real projects. Start anywhere.

Open the first tutorial

Start with your work

Turn one real task into a learning path.

Describe the outcome you need. We’ll match it with tutorials you can use immediately.
Build a useful path

What do you need AI to help you finish?

Bring one real piece of work. We’ll turn it into a small learning path you can use immediately.

Kept on this device. Don’t paste confidential data.

0/280

Browse by topic

Tutorial collections

Jump to a topic, scan every tutorial, and open the guide that answers the question you have now.

Topic 01

Grok Bot

Build a supervised AI coworker system one safe workflow at a time.

0 of 14 complete160 min total
Open first tutorial
  1. T01Paper diorama of three work roles sharing one cloud computer while a human holds an approval gate.
    Not started

    What Is Grok Bot?

    Understand what Grok Bot is, what its named Bots can do, and the shared-computer detail every new user should know.

    8 min
  2. T02Paper diorama of a verified installer branching toward Mac and Windows computers and a first document test.
    Not started

    Grok Bot Pricing, Availability, and Setup

    Check whether your plan includes Grok Bot, install the right desktop app, sign in, and reach your first safe task.

    9 min
  3. T03Paper diorama of one focused research workspace turning three sources into a checked report behind an approval barrier.
    Not started

    Create Your First Grok Bot Agent

    Create one focused Bot, give it a useful job and approval boundary, then test and improve its work.

    9 min
  4. T04Paper diorama of loose work notes passing through a reverse-question funnel into three ranked task cards.
    Not started

    The Grok Bot Brain-Dump Method

    Give a Bot useful context, ask it what to delegate, and turn its suggestions into a safe first automation backlog.

    8 min
  5. T05Paper cutaway of three agent screens wired to one shared computer base as a human takes control.
    Not started

    How Grok Bot's Cloud Computer Works

    Watch, take over, and safely manage the one persistent cloud computer shared by all Bots on your account.

    10 min
  6. T06Paper diorama of a hand demonstrating a browser sequence that winds into a reusable instruction reel.
    Not started

    Teach Grok Bot a Repeatable Task

    Show one Bot a browser workflow, turn the demonstration into a reusable skill, and test it before you put it on a schedule.

    12 min
  7. T07Paper diorama of research, writing, and review stations moving one artifact toward a human approval gate.
    Not started

    Design a Small Grok Bot Team

    Give each Bot a clear job, add specialists only when the handoff is stable, and keep one human approval boundary across the team.

    11 min
  8. T08Paper diorama of a linked evidence folder handed from researcher to writer to reviewer.
    Not started

    Let Grok Bots Hand Work to Each Other

    Use direct handoffs and group chats so Bots can share context without making you copy notes between every stage.

    11 min
  9. T09Paper diorama of a narrow keyed connector retrieving records through a permission gate with revocation shown beside it.
    Not started

    Connect Grok Bot to Apps with Plugins

    Add one connector through Grok Bot's Plugins settings, grant the minimum access, and prove the connection with a read-only task.

    12 min
  10. T10Paper clockwork calendar running a checked skill cartridge through history and approval gates.
    Not started

    Schedule Reliable Work with Grok Bot Routines

    Turn one proven skill into a routine, run a real test, inspect its history, and keep risky actions behind approval.

    13 min
  11. T11Paper diorama of a smartphone monitoring and briefly taking control of a remote cloud computer.
    Not started

    Grok Bot on iPhone: Monitor Work and Take Control

    Use Grok Bot on iPhone to start work, review results, handle approvals, and take over the shared cloud computer safely.

    11 min
  12. T12Top-down paper security map protecting one shared computer with secret, permission, and approval boundaries.
    Not started

    Grok Bot Security, Privacy, and Account Safety

    Protect logins, files, and connected accounts by treating Grok Bot's shared cloud computer as one security zone.

    14 min
  13. T13Paper production line turning one video into source notes, a checked draft, and an article stopped at human approval.
    Not started

    Build a Grok Bot Content Production System

    Build a source-linked workflow that turns one approved video into a researched article draft without automating away editorial review.

    16 min
  14. T14Paper triptych comparing a managed shared cloud workspace, an open self-hosted server, and a multi-channel gateway.
    Not started

    Grok Bot vs Hermes Agent vs OpenClaw

    Choose between a managed Grok Bot workspace and the self-hosted Hermes Agent or OpenClaw using current official documentation.

    16 min

Topic 02

Hermes Agent

Start here if you want an AI assistant that can actually do work.

0 of 12 complete129 min total
Open first tutorial
  1. T01Hand-cut paper Hermes harness coordinating a model core, memory drawer, tool tray, clock, and delivered brief
    Not started

    Hermes Agent Explained

    Learn how Hermes Agent combines memory, tools, skills, schedules, and messaging, and how it differs from chatbots and coding agents.

    7 min
  2. T02Hand-cut paper installation path linking a laptop to a remote server through tool setup, security, and verification gates
    Not started

    Install Hermes Agent on Mac or VPS

    Install Hermes Agent on macOS, Linux, WSL2, Windows, or a VPS, connect one model provider, and verify it with `hermes doctor`.

    8 min
  3. T03Hand-cut paper terminal and desktop workstations connected to one shared Hermes session and memory spool
    Not started

    Hermes Agent TUI vs Desktop App

    Choose between the Hermes TUI and Desktop App for each task. They run the same agent and share state, so you can also use both.

    10 min
  4. T04Hand-cut paper bundle of five reusable instruction cards driving a repeatable research workflow
    Not started

    Hermes Agent Skills and Skill Bundles

    Turn five repeated prompts into reusable Hermes Skills, then group them into a Bundle that runs from one slash command.

    9 min
  5. T05Hand-cut paper MCP catalog bridge connecting Hermes to an external task board through authentication keys and permission controls
    Not started

    Connect Hermes to Linear, GitHub, and Notion with MCP

    Connect Hermes to outside apps through MCP, decide between catalog and custom servers, and install Linear without breaking your setup.

    9 min
  6. T06Hand-cut paper model-routing switchyard sending task cards to efficient engines under a capped coin meter
    Not started

    Choose a Cost-Efficient Model for Hermes

    Read AI model pricing, route routine tasks to lower-cost models, and set a $20 Hermes budget cap on day one.

    11 min
  7. T07Hand-cut paper Hermes server and laptop protected inside a private tunnel while public scanners are blocked outside
    Not started

    Secure Hermes Agent with Tailscale and Scoped Keys

    Secure a Hermes VPS before connecting a paid model account with Tailscale, scoped keys, and a 20-minute setup checklist.

    10 min
  8. T08Hand-cut paper dawn automation workshop gathering research, producing a logged morning brief, and delivering it while the user sleeps
    Not started

    Automate Hermes with Cron, Kanban, and Subagents

    Automate Hermes with daily cron jobs, multi-agent Kanban boards, and sub-agents, then troubleshoot the failures that derail each pattern.

    13 min
  9. T09Hand-cut paper workload case sorting coding, automation, multi-agent, and cloud tasks among four distinct harness workstations
    Not started

    Hermes vs Claude Code, Mavis, and Gemini

    Compare four AI agent harnesses shipping in 2026 by workload, cost, setup, coding, and long-running automation.

    12 min
  10. T10Hand-cut paper always-on VPS cabinet with a containerized Hermes workspace, private laptop access, clock, backup, and budget controls
    Not started

    Run Hermes Agent on a Hostinger VPS

    Choose between Hostinger's self-hosted Hermes VPS and Managed Hermes, compare plans, deploy safely, and budget beyond the server price.

    14 min
  11. T11Hand-cut paper workflow routing a request through named capabilities, active execution, and a separate human verification gate
    Not started

    Install Oh My Hermes and Run a Verified Workflow

    Add Oh My Hermes to an existing Hermes Agent setup, verify the local integration, run a first evidence-aware workflow, and recover cleanly.

    16 min
  12. T12Paper-cutout illustration of an empty window frame like a puppet theater, with a paper cursor pulling levers connected to gears
    Not started

    Set Up Hermes Agent Browser Use Mode Safely

    Enable Hermes Browser Use, choose a local or cloud backend, protect private URLs and credentials, and compare it with the built-in browser tools on one real task.

    10 min

Topic 03

AI Models

Pick the right model for the job without drowning in benchmarks.

0 of 5 complete47 min total
Open first tutorial
  1. T01A hand routes paper task parcels through four stepped levels of a mechanical model-tier tower.
    Not started

    AI Model Tiers: How to Compare LLMs

    Choose an AI model by workload, not leaderboard rank. This four-tier framework helps you find the lowest-cost model that can do the job.

    9 min
  2. T02Three linked paper gears show a fast small model, balanced workhorse, and powerful slow model in one family.
    Not started

    Claude Haiku vs Sonnet vs Opus

    Compare Claude Haiku, Sonnet, and Opus in 2026, then match each tier to the work instead of wasting hundreds of dollars a month.

    9 min
  3. T03A hand installs one of six removable model-weight cartridges into a compact local computer.
    Not started

    Chinese Open-Weight Models: DeepSeek, Kimi, Qwen, GLM, and More

    Compare six Chinese AI labs by workload, and learn how open-weight models differ from open-source software and free API access.

    11 min
  4. T04Three compact paper machines process stacks of routine tasks on a shared conveyor beside a small coin tray.
    Not started

    Qwen, MiniMax, and GLM for Cost-Efficient AI Work

    Qwen, MiniMax, and GLM handle 80% of daily AI work at 5 to 20 times lower cost than Claude Opus 4.8. Choose by workload.

    9 min
  5. T05A task parcel reaches a railway switch between a remote compute tower and a local open-weight workshop.
    Not started

    Frontier vs Open-Weight Models: A Decision Guide

    Combine L01 through L04 into five questions that choose an AI model for a task and show when to switch tiers mid-project.

    9 min

Topic 04

Case studies

Move from watching tools to shipping something useful.

0 of 12 complete144 min total
Open first tutorial
  1. T01A paper coding arm edits repository file cards as failed and passing code strips circle a test loop.
    Not started

    Claude Code for Beginners: Build with a Coding Agent

    Put Claude Code inside a real repository and use a write, test, fix loop to build safely. See when Hermes or Cursor is a better fit.

    9 min
  2. T02A taped cardboard prototype passes through an inspection rig and becomes a braced production mechanism.
    Not started

    Vibe Coding vs Real Coding

    Use vibe coding for disposable prototypes and real coding for work that must last. A 2x2 framework shows when to switch.

    8 min
  3. T03A folded project plan passes from a research desk to a coding and test workbench for review.
    Not started

    Build a Side Project with Hermes and Claude Code

    See what one person can ship in a weekend when Hermes handles research and the specification, then Claude Code builds and iterates.

    9 min
  4. T04A visitor compares five verified destination cards on a paper map and follows one selected route.
    Not started

    Build a Niche Directory Site with AI

    Validate a niche directory with five useful records, real user decisions, and buyer interviews before scaling SEO, data, or monetization.

    12 min
  5. T05A manual customer delivery activates a layered small-business system while an automation arm waits for approval.
    Not started

    A Practical AI Stack for a Small Online Business

    Find one painful customer problem and deliver the offer manually. Add a landing page, payments, email, analytics, and automation as the work repeats.

    11 min
  6. T06Hundreds of article cards cross a locked bridge from a tangled CMS machine into clean static page shelves.
    Not started

    We Ditched WordPress for Astro and Saved Thousands

    Boxmining moved 544 posts from WordPress to Astro 7 and Vercel after plugin bloat, slow pages, and four-figure hosting bills.

    8 min
  7. T07Hand-cut paper control room with one lead operator connected to three coding-agent workstations and a reconnecting laptop
    Not started

    Run a Multi-Agent Coding Team with Herdr

    Install Herdr, keep coding agents in persistent terminals, and coordinate isolated workers from one orchestrator without confusing status with proof.

    20 min
  8. T08Hand-cut paper dispatch tower routing three coding agents into separate Git workspaces that converge at a review gate
    Not started

    Gas Town: Give Every Coding Agent Its Own Workspace

    Run several coding agents without putting them in the same checkout: track work, create separate Git worktrees, monitor progress, and recover safely.

    16 min
  9. T09Hand-cut paper terminal operator monitoring five parallel agent lanes and reconnecting one blocked session to a shared completion rail
    Not started

    Warp for Multi-Agent CLI Work: A Practical Team Setup

    Turn Warp into a clear operator cockpit for parallel CLI agents using vertical tabs, Git worktrees, repeatable Tab Configs, notifications, and a final verification pass.

    16 min
  10. T10Editorial diagram of independent workflows converging into one verified result
    Not started

    Build and Verify a Claude Code Dynamic Workflow

    Turn a large Claude Code job into independent workers, a deliberate merge, and a separate verification pass without creating an expensive agent swarm.

    9 min
  11. T11Live Boxmining homepage after its migration to Cloudflare Pages
    Not started

    Migrate a Static Site from Vercel to Cloudflare Pages

    Move a static production site through a preview deployment, canonical-domain cutover, redirect audit, cache check, and rollback-safe verification.

    11 min
  12. T12Hand-cut paper coding-agent harness routing a repository file through removable tool modules, a human approval switch, and a test loop
    Not started

    DeepSeek Harness Tutorial: Run Your First Coding Agent Safely

    Install DeepSeek Harness, connect a model, choose a workspace, run one controlled coding task, and verify every change before exploring its plugin system.

    15 min