How Sam, The Aspiring AI Builder, Uses WayOfTeams

A step-by-step walkthrough of how a curious, hands-on beginner moves from "chat with a bot" to building real working utilities — using WayOfTeams as the bridge between basic AI chat and actual agentic workflows.

Who Sam Is

Sam uses ChatGPT through the browser. He asks it to write emails, summarize documents, explain code he found online. He's seen YouTube videos about "AI agents" and "MCP" and wants in — but every guide he finds assumes he already knows what a terminal is, what an API key does, or why context windows matter.

He doesn't need another framework. He needs a clear path from "I type prompts into a box" to "I built something that actually works."

Phase 1: Signing Up (Minutes 0–3)

What Sam Does

He goes to teams.zerwiz.org, clicks Sign In, and signs up with GitHub (or Google, or Apple). One click. Done.

What He Sees

The onboarding flow. It asks him one question:

"What do you want to build?"

This isn't a setup wizard for infrastructure. It's the start of a conversation. Sam types something like:

"I want a tool that monitors my email and sends me a daily summary of what matters."

What Happens Next

WayOfTeams captures his answer as a structured problem statement. It asks follow-up questions:

Who is this for?
What does success look like?
What should it be called?

Sam answers in plain language. No syntax. No config files. No terminal.

Why This Works for Sam

He's been overwhelmed by every AI setup guide he's tried. This one starts with a question he can answer. The system does the rest.

Phase 2: The Dashboard — His Command Center (Minutes 3–10)

What Sam Sees

A dark-themed, single-page dashboard. Everything one click away.

Sidebar:
  📊  Dashboard        📋  Standups
  🎫  Tickets          🤖  AI Chat
  📋  Kanban           💬  Team Chat
  📊  Skills           💡  Ideas
  ✅  Review Queue     📰  News
  👥  Developers       🔧  Admin
  ⚙  Settings         👤  My View
  📄  Docs             🔗  F-rr-d Changes

He doesn't know what half of these are yet. That's fine. He sees:

Stats: recent tickets, today's standup, active skills
AI Chat panel: embedded right in the dashboard — no separate tab
Recent activity: what's been created, what's in progress

What He Clicks First

AI Chat. It's familiar. It looks like the chat interfaces he already uses. But this one is connected to his project.

Why This Works for Sam

The dashboard doesn't dump him into a blank terminal. It shows him what exists, what's happening, and gives him one clear next step: talk to the AI.

Phase 3: His First Conversation with AI (Minutes 10–20)

What Sam Does

He opens the AI Chat panel and types:

"I want to build a tool that monitors my email and sends me a daily summary. Can you help me figure out what I need?"

What the AI Does

The AI doesn't give him a wall of code. It asks questions:

What email provider do you use?
How do you want to receive the summary? (email, Telegram, dashboard?)
Do you want it to filter by sender, subject, or importance?

Sam answers each one. The AI is building context — the same context that would normally be lost if he closed the browser tab.

What Happens Behind the Scenes

The AI creates a ticket automatically:

Title: "Build email monitoring tool with daily summary"
Status: "Backlog"
Category: "Feature"

It also creates a kanban card on his board. Sam can see both in the dashboard immediately.

What Sam Doesn't Know Yet

The AI stored his preferences in Anchor memory — next session, it remembers his email provider, his preferred summary format, and the decisions they made together
The ticket was pushed to his f-rr-d repo — a Git-backed change log that tracks everything from idea to implementation
The kanban card is real-time — if he had teammates, they'd see it instantly

Why This Works for Sam

He's used ChatGPT where the conversation disappears when he closes the tab. Here, every answer he gives becomes persistent context. The AI remembers. The ticket exists. The card is on the board. Nothing is lost.

Phase 4: The Guided Journey — From Idea to Working Software (Day 1–3)

What Sam Does

He follows the 8-step Way-Of Journey that appears in his onboarding:

1. What do you want to do?
2. What do you want to name it?
3. Connect your AI tool (Teams MCP / harness).
4. Let's make two repos based on the name.
5. Let's make your first ticket.
6. Let's implement your first ticket.
7. Look at what you just did.
8. Let's create your next tickets.

He's already done steps 1–5. Now the AI walks him through step 6:

"Let's implement your first ticket. I'll create a plan, then build it step by step. You review each step before I move on."

What the AI Builds

The AI generates a phased plan:

Phase 1: Set up the email connection (OAuth with Gmail)
Phase 2: Build the filter logic (importance scoring)
Phase 3: Generate the daily summary (format as markdown)
Phase 4: Send via Telegram (or email, or both)

Each phase produces working code. Sam sees it happen:

The AI writes the code
The AI runs tests
The AI commits to Git
The dashboard updates in real-time

What Sam Sees

Ticket: status changed from "Planned" → "In Progress" → "Done"
Kanban: card moved across columns
Git history: clean commits with ticket references
Dashboard: everything connected, everything visible

Why This Works for Sam

He's never seen code get written and committed in front of him like this. The AI didn't just give him a code snippet to copy-paste. It built the thing, tested it, committed it, and showed him the result. He can click through the git history and see exactly what changed and why.

Phase 5: Learning the Vocabulary (Day 3–7)

What Sam Encounters

As he uses WayOfTeams, he starts picking up terms he's seen in YouTube videos but never understood:

Tickets — A structured description of work. Problem statement, desired outcome, acceptance criteria. Not a chat message. A persistent, trackable unit of work.

Plans — A step-by-step implementation guide. The AI reads the ticket, analyzes the codebase, and produces a phased plan with file changes and success criteria.

Kanban — A visual board where cards move from left to right. Each card is a ticket. Moving a card changes the ticket's status. It's real-time.

Standups — A daily summary of what was done, what's planned, and any blockers. Generated automatically from git history and ticket changes. No meeting required.

Skills — Pre-built instructions for specific tasks. 53 of them. Things like "create a plan," "validate a plan," "commit changes," "write tests." The AI uses them automatically — Sam doesn't need to learn them all at once.

The Harness — The AI Engineering Harness. The collection of skills, agents, and commands that power the AI's behavior. It's already installed. Sam just uses it.

How Sam Learns

He doesn't read a manual. He hovers over things in the dashboard and sees tooltips. He asks the AI:

"What's a plan?"

The AI explains in plain language and shows him one. He asks:

"What skills are available?"

The AI lists them with descriptions. He asks:

"How do kanban boards work?"

The AI walks him through creating one.

Why This Works for Sam

He learns by doing, not by reading documentation. WayOfTeams lets him ask questions in context — the AI knows what he's looking at and explains it in relation to his actual project.

Phase 6: The Skills He Uses First (Week 2)

What Sam Discovers

The 53 skills are organized by when you'd use them. Sam gravitates toward the ones that feel immediately useful:

Skill	When Sam Uses It	What It Does
`ticket-manager`	"I need to organize my tasks"	Creates, updates, manages tickets
`create-plan`	"I have a ticket, now what?"	Turns ticket into step-by-step plan
`validate-plan`	"Is this plan solid?"	Checks feasibility before building
`implement-plan`	"Let's build this"	Executes plan phase by phase
`commit`	"Save this work"	Creates well-structured git commits
`standup`	"What did I do today?"	Generates daily standup
`debug`	"Something's broken"	Reads logs, examines state, proposes fix
`help-command`	"What can I do?"	Shows all skills, commands, agents

What He Doesn't Use Yet

The advanced skills — tdd, github-pr, postmortem-manager, observability-driven-development. That's fine. He'll get there. The harness grows with him.

Why This Works for Sam

He's not overwhelmed by 53 skills. He uses the 8 that matter right now. The others are there when he's ready.

Phase 7: Context That Doesn't Leak (Week 2–3)

Sam's Problem

Every AI chat he's had before loses context. He explains his project in paragraphs. The AI responds helpfully. He closes the tab. Next session, the AI has no idea what he's building.

How WayOfTeams Solves It

Three layers of persistent context:

1. Memory (Anchor) Every conversation, decision, and pattern gets stored semantically. When Sam starts a new session, the AI pulls in relevant memories automatically via context_inject.

"What did we decide about the email filter?"

The AI finds the decision from three days ago and reminds him.

2. Knowledge Base Sam stores learned patterns in the project knowledge base:

"Save this: Gmail OAuth tokens expire after 7 days. Refresh tokens don't expire. Always use refresh tokens for long-running tools."

Next time the AI works on email features, it reads this knowledge before planning. It won't suggest an approach that uses short-lived tokens.

3. Rules Sam creates coding rules:

"Always handle errors gracefully. Never let a network request crash the tool. Log errors to a file and send a Telegram alert."

Rules are enforced by agents during implementation. The AI follows them because they're part of the project context, not just a chat message.

What Sam Feels

For the first time, the AI feels like it knows his project. Not because he's pasting the same context every session, but because the context is stored, searched, and injected automatically.

Why This Works for Sam

Context leakage was his biggest frustration with ChatGPT. WayOfTeams solves it with three complementary systems that work behind the scenes. He doesn't need to understand Anchor's architecture — he just notices that the AI remembers.

Phase 8: Kanban Without the Overhead (Week 3)

What Sam Does

He creates his first kanban board. The AI helps:

"Create a board called 'Email Monitor' with columns: Backlog, In Progress, Done"

The board appears. He drags cards across columns. Real-time sync means if he opens the dashboard on his phone, the board looks the same.

He creates a private board (scope: private) for personal tasks. Nobody else in his company sees it. He's the only one with access.

He uses the Today view to see what's due. Reminders pop up for overdue items.

What He Notices

Cards are tickets — the same work shows up in dashboards, standups, and reports
He can attach files to cards — a log dump, a screenshot, a config file
The calendar view shows deadlines alongside his standups
He can filter and search across all his boards

Why This Works for Sam

He's tried Trello, Notion, Todoist. They're all separate from his AI tools. WayOfTeams kanban is connected to the same system his AI uses. When the AI completes a ticket, the card moves automatically. When he drags a card, the ticket status updates. Nothing is manual.

Phase 9: Ideas Worth Capturing (Week 3–4)

What Sam Does

He has ideas at random times. "What if the tool could also check my calendar?" "What if it could draft replies?"

He goes to the Ideas page and submits them. Each idea gets:

A title and description
An upvote/downvote system
A status workflow (Idea → Under Review → Accepted → In Progress → Done)
Linkable tickets

He upvotes his own ideas to prioritize them. When he's ready, he says:

"Turn my top idea into a ticket"

The AI does it. The idea links to the new ticket. The kanban card appears.

Why This Works for Sam

Ideas used to live in his head or in random notes app entries. Now they're in a system that connects directly to the work pipeline. An idea becomes a ticket becomes a plan becomes working software — all in one place.

Phase 10: Growing Into the System (Month 2+)

What Sam Does

He's comfortable now. He starts exploring:

Team Chat — He invites a friend. They message each other, share screens, and the AI captures their conversation as tickets and plans. Meetings become work items, not forgotten notes.

Notifications — He sets up Telegram alerts. When a ticket is created, moved, or completed, he gets a message on his phone. He never misses what his AI built while he was away.

Docs — He creates his first documentation page: "How the Email Monitor Works." The AI helps write it. It lives in the project, searchable and always up to date.

Standups — He runs /standup and sees a structured summary of his day: what was done, what's planned, any blockers. Generated from git history and ticket changes. He didn't write a single word.

Versions — He bumps the version from 0.1.0 to 0.2.0. The AI generates a CHANGELOG entry with categorized changes (Added, Fixed, Changed). It commits the version bump and changelog update automatically.

What He's Building

The email monitor tool is working. He's added features:

Telegram notifications
Importance scoring
Daily digest with AI summaries
A simple web dashboard

He's shipping features weekly. Each feature follows the same loop: ticket → plan → implement → review → learn.

Why This Works for Sam

He's not learning a new tool every week. WayOfTeams is his operating system for building things. The more he uses it, the more context it has, the better the AI performs, the more he builds. It's a flywheel.

Feature Adoption Order

Based on Sam's profile, here's what he'd use and when:

Priority	Feature	Why Sam Cares
1	AI Chat	Familiar interface, like ChatGPT but connected to his project
2	Onboarding Journey	Clear step-by-step path, no infrastructure knowledge needed
3	Tickets	Structured work items that don't disappear when he closes a tab
4	Kanban	Visual board with drag-and-drop, real-time, multiple views
5	Memory/Anchor	AI that remembers his decisions and context across sessions
6	Skills (core 8)	Pre-built instructions the AI uses automatically
7	Standups	Auto-generated daily summaries, no meeting overhead
8	Ideas	Capture ideas, vote, link to tickets
9	Knowledge + Rules	Store patterns and enforce conventions
10	Docs	Living documentation that stays current
11	Notifications	Telegram alerts for project activity
12	Versions	Automated SemVer + CHANGELOG

What Makes WayOfTeams Different for Sam

Problem Sam Has	WayOfTeams Solution
"AI forgets everything between sessions"	Anchor memory + Knowledge + context injection — AI remembers automatically
"I don't know where to start"	8-step guided journey with one question at a time
"Every guide assumes I know terminals and APIs"	Web-based dashboard, AI chat, no terminal required to start
"I can't see what my AI built"	Dashboard shows every ticket, commit, plan, and session
"My ideas disappear"	Ideas page with voting, status workflow, and ticket linking
"Documentation is always outdated"	Living docs tied to the same system that does the work
"I don't know what skills to use"	53 skills, but the AI picks the right ones automatically
"I want to build real things, not just chat"	Ticket → Plan → Implement → Review → Ship loop

The Core Loop

For Sam, WayOfTeams boils down to one simple progression:

  Idea → Ticket → Plan → Build → Ship → Learn
   ↑                                        │
   └────────────────────────────────────────┘

Every iteration makes the system smarter:

Ideas capture what he wants to build
Tickets describe the work clearly
Plans tell the AI how to build it
Building produces working software with clean commits
Shipping makes it real
Learning (memory, knowledge, rules) prevents the same mistakes twice

The loop runs through a dashboard he understands, AI he can talk to in plain language, and tools that grow with him.

That's the system Sam would actually use.

← Back to all personas