Software Trainings · Agentic Coding track
Agentic coding — shipping with AI as a teammate.
A six-week track in disciplined AI-pair-programming: Cursor, Claude Code, Copilot, custom MCP servers, and multi-agent planner/coder/reviewer loops applied to your own codebases — not demos.
What this track delivers
Why this track is different.
- 01
Fluent agent IDEs
Cursor and Claude Code workflows: rules files, project context, multi-file edits, long agent runs.
- 02
Build an MCP server
Expose your tools — docs, DB, deploy, observability — to coding agents through a custom MCP server.
- 03
Multi-agent pipelines
Planner / coder / reviewer loops with queues, budgets, hand-offs and human checkpoints.
- 04
Review & safety
Tests-as-guardrails, diff review, rollback strategy, and secret hygiene for AI-generated code.
Week-by-week
Syllabus in order.
- Week 1AI-pair foundations
Context windows, file selection, when to drive vs let the agent drive.
- Week 2Cursor & Claude Code
Rules files, project context, multi-file edits, agent runs.
- Week 3MCP servers
Building tools agents can call — docs, DB, deploy, observability.
- Week 4Multi-agent workflows
Planner / coder / reviewer loops, queues, budgets, hand-offs.
- Week 5Review & safety
Diff review, tests-as-guardrails, rollback, secret hygiene.
- Week 6Capstone
Ship a real feature in a real repo using an agentic pipeline.