Skip to content

Appendix F — Instructor Run-of-Show (Minute-by-Minute)

All times are local. Morning Theory starts at 9:00 AM and afternoon Labs start at 2:00 PM.


Day-1 AM — Theory (4h) · 9:00–13:00

Time Segment Cues & Expected State
09:00–09:10 Welcome & goals Outcomes, repo map, docs open
09:10–09:30 3.1 What is MCP and Why a Gateway Pain points; UI concepts; value
09:30–10:20 3.2 Architecture and Core Concepts Flow; plugin hooks; multitenancy
10:20–10:30 Break
10:30–11:35 3.3 Agents & Clients Overview CrewAI vs LangGraph; Demo MCP CLI
11:35–12:10 3.4 Serving Patterns: gRPC, REST, Wrappers curl passthrough demo
12:10–12:20 Break
12:20–12:50 3.5 Security & Governance Essentials RBAC, OAuth/mTLS, teams
12:50–13:00 3.6 Observability & Telemetry Log fields & IDs; OpenTelemetry

Day-1 PM — Labs (4h) · 14:00–18:00

Time Segment Cues Outputs
14:00–14:20 4.1 Lab 0: Environment Checks Verify env; unblock ports Checklist complete
14:20–15:05 4.2 Lab 1: Quickstart (Gateway Up + Well-Known) mcpgateway up; /health Health + tools screenshots
15:05–15:15 Break
15:15–16:00 4.3 Lab 2: Register Your First MCP Server Register server; invoke Tool appears; valid output
16:00–16:35 4.4 Lab 3: Clients & CLI Flows Two clients call same tool 2 screenshots
16:35–16:45 Break
16:45–17:20 4.5 Lab 4: Simple Passthrough / Wrapper Wrapper/passthrough demo JSON response
17:20–18:00 4.6 Lab 5: Guardrails (Rate Limit) Enable rate limit; show 429 429 + 200 proofs

Day-2 AM — Theory (4h) · 9:00–13:00

Time Segment Cues State
09:00–09:10 5.1 Capstone Overview & Success Criteria Success checklist Clear target
09:10–09:25 5.2 Introduction & Architecture Diagram Langflow↔Adapter↔Gateway↔Agent Shared model
09:25–09:40 5.3 Setup & Prerequisites Dry-run commands Ready to start
09:40–10:20 5.4 Designing the Langflow Tool (I/O and API) Define I/O schema; talk latency Flow chosen
10:20–10:30 Break
10:30–11:10 5.5 Exposing Langflow as an MCP Tool Server Show adapter skeleton & registration Endpoints known
11:10–12:00 5.6 The CrewAI Agent Pattern Minimal code; invoke via gateway Code plan ready
12:00–12:10 Break
12:10–12:50 5.7 Hardening: Guardrails, RBAC, (OBO), Logs Pick policies + RBAC; optional OBO Acceptance plan
12:50–13:00 Q&A Clarify open items Confidence high

Day-2 PM — Labs (4h) · 14:00–18:00

Time Segment Cues Outputs
14:00–14:15 6.1 Lab A: Setup & Prereqs Compose up; venv; install deps Ready checklist
14:15–14:55 6.2 Lab B: Build the Langflow Tool Endpoint responds Sample JSON
14:55–15:25 6.3 Lab C: Expose as MCP Tool Server Register adapter; catalog visible Tool + CLI proof
15:25–15:35 Break
15:35–16:10 6.4 Lab D: Build the CrewAI Agent Run agent; confirm tool call Agent output
16:10–16:40 6.5 Lab E: Guardrails in Action Show 429 + redaction Policy proofs
16:40–16:50 Break
16:50–17:20 6.6 Lab F: RBAC (+ Optional OBO) 200 vs 403; optional OBO Authz + logs
17:20–17:40 6.7 Lab G: Observability Trace Correlation ID + latency 3-line trace
17:40–18:00 6.8 Team Demos & Rubric 3–4 min/team; rubric scoring All proofs in

← Back to Syllabus