Build Tekton — a multi-cloud, multi-observability AI platform — end-to-end. From a single CLI to a knowledge-graph brain to an AI SRE in Slack to a cross-service code review agent.

Summary

Phase	Title	Modules	Approx classes (2h each)
0	Foundations	8	6–8
1	Tekton CLI v1 (Datadog + AWS)	10	9–10
2	Tekton CLI v2 (K8s data plane)	8	8
3	Tekton CLI v3 (MCP + Enterprise RCA)	8	8
4	Infra Brain (Correlated Knowledge Graph)	13	12–14
5	AI SRE Slack Bot	8	8
6	Cross-Service Code Review Agent	8	7–8
—	Cross-cutting tracks (Security, Cost, Reliability, Evaluation)	4	woven into above
Total		63 modules	~58–64 classes

What you will build !

By the final class, you will have shipped:

Tekton CLI — a multi-cloud, multi-observability CLI that any AI agent (Claude, Cursor, Codex) can use as a tool.
K8s data plane — agent + gateway that streams cluster events from any cluster to your own raw store (Quickwit / OpenObserve).
MCP servers — Datadog and OpenObserve / Quickwit exposed to AI agents as tools, with pre-aggregation so you don't burn tokens.
Tekton Correlation Engine — joins traces, K8s events, deploys, code commits, and cloud events on shared identifiers.
Infra Brain — a UI-driven connector platform that ingests every data source into a single queryable knowledge graph.
AI SRE Slack Bot — conversational interface to the brain; gives full RCA in seconds inside Slack threads.
Cross-Service Code Review Agent — reviews PRs with full infrastructure + downstream-service context.

Summary

What you will build !

PART 1 — Module List