Skip to content
Jordan Hudgens

Dev Guides

Technical training you can run, not just read.

Each guide pairs a written lesson with hands-on labs that fail on purpose, an anchored assessment and the material to teach it to a room. Companion videos and short clips accompany the guides that have them.

guides
12
hands-on labs
37
AI-generated illustration

Featured · Guide 01 · Intermediate

Build a bounded agent loop in TypeScript, break it with seeded faults, and repair it so the system of record, not the model's message, decides success.

  • Build a bounded tool-use loop
  • Trace every call to its result
  • Diagnose a seeded failure
Start the guide40 min read · 90 min workshop
  • Written guide
  • Labs
  • Instructor kit
  • Downloads

Find a guide by skill

Filter by skill, level and format. Guides list only what they include.

12 of 12 guides

  • Guide 01Intermediate
    AI-generated illustration

    Agents and tool use

    Build, Break, and Repair a Reliable Tool-Using Agent

    Build a bounded agent loop in TypeScript, break it with seeded faults, and repair it so the system of record, not the model's message, decides success.

    • Written guide
    • Labs
    • Instructor kit
    • Downloads
    40 min read · 3 labs
  • Guide 02Intermediate
    AI-generated illustration

    Evaluation and testing

    Agent Evaluations That Measure Real Outcomes

    Grade what an agent left in the system of record, run every task more than once, and turn the numbers into a ship or hold decision you can defend.

    • Written guide
    • Labs
    • Instructor kit
    • Downloads
    60 min read · 3 labs
  • Guide 04Intermediate
    AI-generated illustration

    MCP and integrations

    Build and Debug Secure MCP Integrations

    Put company services behind one version-pinned MCP server that several AI clients share, prove it refuses the wrong callers, and debug it from the bytes up.

    • Written guide
    • Labs
    • Instructor kit
    • Downloads
    63 min read · 3 labs
  • Guide 05Intermediate
    AI-generated illustration

    Claude Code and agentic coding

    Claude Code Workflows You Can Verify

    Freeze the spec before the agent touches code, grant only what the task needs, and accept a change only on evidence the change could not edit.

    • Written guide
    • Labs
    • Instructor kit
    • Downloads
    75 min read · 3 labs
  • Guide 06Intermediate
    AI-generated illustration

    Instructor development and assessment

    From Expert to Cleared-to-Teach: Instructor Calibration

    Clear an instructor on observed evidence: score a teach-back with anchors, calibrate two observers, and scope the record to one module version.

    • Written guide
    • Labs
    • Instructor kit
    • Downloads
    59 min read · 4 labs
  • Guide 07Intermediate
    AI-generated illustration

    Claude Code and agentic coding

    Route Claude Code Work Across Model Tiers and Prove the Savings

    Put search on a small model and edits on a mid-tier one, confirm in telemetry which model ran each request, and keep the route only if checked runs spend less.

    • Written guide
    • Labs
    • Instructor kit
    • Downloads
    69 min read · 3 labs
  • Guide 08Intermediate
    AI-generated illustration

    Context and retrieval

    Context Engineering for Long-Running Agents

    Run one support shift under three context strategies on the Claude API, grade every ticket, and keep a strategy only when its outcomes hold.

    • Written guide
    • Labs
    • Instructor kit
    • Downloads
    96 min read · 3 labs
  • Guide 09Advanced
    AI-generated illustration

    Enterprise deployment

    Enterprise AI Deployment: Scope, Secure, Prove, and Hand Over

    Take a request to "automate support" to a narrow pilot with a measured baseline, tested decisions, a security review and a handover that runs.

    • Written guide
    • Labs
    • Instructor kit
    • Downloads
    102 min read · 3 labs
  • Guide 10Intermediate
    AI-generated illustration

    Claude Code and agentic coding

    Agent Skills: Package Expertise into Tested Workflows

    Turn a pasted review prompt into a portable Agent Skill with progressive loading, stream-graded activation, baseline record tests, and clear trust boundaries.

    • Written guide
    • Labs
    • Instructor kit
    • Downloads
    74 min read · 3 labs
  • Guide 11Intermediate
    AI-generated illustration

    Agents and tool use

    Design Tool Contracts Agents Can Use Correctly

    Turn a confusing tool library into contracts a model reads correctly, then prove the difference on the same tasks with the system of record, not the reply.

    • Written guide
    • Labs
    • Instructor kit
    • Downloads
    69 min read · 3 labs
  • Guide 12Intermediate
    AI-generated illustration

    Claude Code and agentic coding

    Extend Claude Code from the Inside: Build a Forkable Visual Mod

    Build Model Ledger, a Claude Code mod that draws per-model tokens and estimated cost in the session, test it offline, break it, and decide which mods may load.

    • Written guide
    • Labs
    • Instructor kit
    • Downloads
    88 min read · 3 labs

How every guide is built

Evidence before claims.

  1. 01

    Runnable

    Every lab is executed before publishing, with exact versions, commands and outputs recorded.

  2. 02

    Broken on purpose

    Each guide includes a seeded failure, its symptom, the diagnosis and how to verify the fix.

  3. 03

    Assessed

    Objectives map to a practical task and an anchored rubric, not a recall quiz.

  4. 04

    Teachable

    An instructor kit covers run of show, 12 and 100 learner formats, dry runs and clearance.