GOAL WRITE RUN INSPECT VERIFY HARDEN AUTOMATE
The course map.
A practical path from your first Kujo program to bounded AI-native systems.
Your learning path
40 lessons · 6 stage builds · 1 capstone
01
Kujo foundations
Learn · practice · verify
- 01What Kujo is
- 02Installation, CLI, and your first program
- 03Bindings, values, and mutability
- 04Values, operators, and truthiness
- 05Control flow and block scopes
- 06Functions, callbacks, and return values
- 07Collections, pipes, and null handling
02
Language semantics
Learn · practice · verify
- 08Scope, shadowing, and capture
- 09Structs and enums
- 10Match, Result, and Option
- 11Exceptions and recoverable errors
- 12Optional typing without imaginary guarantees
- 13Async functions and await
- 14Spawn and bounded concurrent work
03
Projects and tooling
Learn · practice · verify
- 15Modules and explicit exports
- 16Projects, packages, and lockfiles
- 17Finding the standard library contract
- 18Testing, fixtures, and runtime evidence
- 19Diagnostics and machine-readable interfaces
- 20Documentation as a build artifact
- 21Editor integration and the LSP boundary
04
Native automation
Learn · practice · verify
- 22Filesystem automation with bounded paths
- 23Processes without unnecessary shells
- 24HTTP and network boundaries
- 25Structured data and local persistence
- 26Trusted execution and minimum authority
- 27Secrets, redaction, and endpoint policy
05
AI-native programming
Learn · practice · verify
- 28AI effects and portable messages
- 29Deterministic request identity
- 30Record, replay, and offline evidence
- 31Streaming and cancellation
- 32Structured outputs and schema validation
- 33Context budgets, embeddings, and vectors
- 34Bounded AI tool loops
06
Agentic systems
Learn · practice · verify
- 35Mechanism versus policy
- 36Loop Engineering, specs, and stop conditions
- 37Evaluation independent of generation
- 38Evidence, run records, and work boundaries
- 39Isolation, observability, and watchdogs
- 40MCP and external tool interoperability
FINAL CAPSTONE
Local-first AI operations runner.
Validate a request, evaluate the result, and record the evidence.