aGeNT1.01
Built.Maintained.Explained.
The first course in agentic AI. Taught by a website that runs on it. What you learn here is what keeps this site alive.
synced_2_items_to_sanity
12h ago
- PR #40 mergedsuccess
VideoHunter: 2026-04-15-0851 — 2 tutorials staged
12h ago
- ContentScoutsuccess
synced_5_items_to_sanity
12h ago
- PR #39 mergedsuccess
Content Discovery: Agent Orchestration & Harness Patterns (2026-04-15)
12h ago
- VideoHuntersuccess
video-discovery: 2 tutorials staged
14h ago
- ContentScoutsuccess
content-discovery: reviewed 152 candidates from 8 sources, staged 5 items (all score ≥8) covering agent orchestration, deployment, and harness architecture patterns
15h ago
- VideoHuntersuccess
video-discovery: 0 candidates fetched
1d ago
- ContentScoutsuccess
synced_5_items_to_sanity
1d ago
- PR #38 mergedsuccess
ContentScout: Stage 5 high-value agent architecture & deployment items
1d ago
- PR #37 mergedsuccess
VideoHunter: 2026-04-14-1014 — 6 tutorials staged
1d ago
- PR #36 mergedsuccess
ContentScout: 2026-04-14-0910 content discovery
1d ago
- PR #35 mergedsuccess
VideoHunter: 2026-04-14-0821 — 6 tutorials staged
1d ago
Site Mission
The website
that explains
itself.
An agentic AI showcase built by its owner, maintained by its agents. Every page presents curated content about agentic AI — and those same agentic principles keep the site alive.
The Concept
The course and the case study, in one site
Agent 1.01 — as in the first course in agentic AI — is simultaneously the product and the proof. Every page presents curated content about agentic AI systems: agents, protocols, architecture patterns, tutorials, and more. And the very same agentic principles described on those pages are what keep the site alive, updated, and growing.
The site doesn't just talk about agentic AI — it is agentic AI in action. You're reading content that was discovered, written, reviewed, and published by AI agents running on a schedule. The agents open pull requests; a human reviews and merges them. That's the whole loop.
The Loop
Who Built It
One human. Eight agents.
The site was designed and built by a single developer who wanted to understand, teach, and demonstrate agentic AI by actually building with it. The infrastructure, design system, and agent architecture were all hand-crafted.
After launch, a team of eight specialised agents took over ongoing maintenance. Each has a specific trigger, function, and scope — no overlap, no ambiguity.
ContentScout
Daily
Discovers new resources & opens PRs
QualityGuard
PR event
Reviews every PR before it reaches the owner
VideoHunter
Daily
Finds relevant tutorials on YouTube
ContentWriter
Issue
Drafts content summaries and roundups
LinkDoctor
Weekly
Checks for broken links & opens fix PRs
SEOOptimizer
Weekly
Audits and improves meta & OG tags
DependencyBot
Weekly
Keeps npm dependencies up to date
SiteBuilder
Issue
Implements new features from GitHub issues
Why
Learning by doing — in public
Most writing about agentic AI is abstract. Agent 1.01 is an attempt to make it concrete: every agent is documented with its trigger, function, and last-run status. Every piece of content carries the name of the agent that found it. The cost of running the agents is displayed openly.
“The goal isn't perfection — it's transparency.”
If an agent makes a mistake, that's part of the story. If a new protocol emerges overnight and ContentScout finds it by morning, that's also the story.