Course Objective & Scope
Shipping a net-new vertical feature, restructuring coupled modules (refactoring) and upgrading a framework or SDK across hundreds of files (migration) require fundamentally different agent strategies. This course provides concrete, battle-tested execution playbooks for each of these three core engineering workloads.
Participants practise golden-slice feature implementation across DB/API/UI/tests, staged behaviour-preserving multi-file refactoring, and hybrid migrations where coding agents write deterministic AST codemods (jscodeshift, GritQL, LibCST, OpenRewrite) for the 80% bulk and resolve the remaining 20% edge cases in closed test loops.
Who Should Attend?
- Backend, frontend, full-stack and mobile software engineers
- Tech leads and platform/infrastructure engineers
- Teams executing framework upgrades, SDK migrations or architectural refactoring
- Prerequisite: experience with Git, automated testing and basic coding agent workflows
Key Takeaways
- Deliver full-stack features across DB, API, UI and tests by anchoring the agent to a canonical repo slice
- Sequence multi-file refactorings into buildable, behaviour-preserving commits
- Have the agent author deterministic AST codemods for repetitive changes across hundreds of files
- Finish complex migration edge cases using compiler- and test-driven agent loops
- Ship clean, easily reviewable stacked PRs for features, refactors and migrations
Curriculum
Three Workloads, Three Distinct Agent Strategies
- Contrasting additive feature work, invariant-preserving refactors and fleet-wide migrations
- Scoping risk, context requirements and verification gates per workload
The Feature Playbook: Golden-Slice End-to-End Delivery
- Anchoring DB schema, service logic, API routes, UI components and tests to a reference slice
- Feature flags and splitting delivery into 2–3 reviewable stacked PRs
The Refactor Playbook: Safe Multi-File Restructuring
- Locking behaviour with tests, introducing adapters/interfaces and migrating callers incrementally
- Combining deterministic LSP symbol moves with agent-driven logic restructuring
The Migration Playbook: Hybrid Codemods + Agent Loops
- Why you should not edit 500 files with raw LLM calls: having the agent write AST codemods first
- Using closed-loop agent passes to resolve the remaining 15–20% non-trivial edge cases
Dual-Version Bridging, Verification and PR Packaging
- Adapter patterns and incremental rollout so the main branch stays green throughout
- Structuring PR narratives, verification evidence and rollback plans for fast review
Other Courses in This Category
-
Hands-on training in autonomous QA workflows using Playwright MCP, browser-driving testing agents, self-healing E2E tests and autonomous exploratory testing across UI and API user journeys.
- Live Course
- 5 modules
-
Solving the PR review bottleneck in AI-accelerated teams through stacked diffs, low-noise AI review bots, evidence-backed PR templates and risk-tiered human approval workflows.
- Live Course
- 5 modules
-
Advanced training on running autonomous coding agents safely on developer machines and servers using container/microVM sandboxes, filesystem/network egress controls, least-privilege MCP permissions and secret isolation.
- Live Course
- 5 modules
-
Scaling AI coding from solo tricks to team workflows: building version-controlled agent skills (SKILL.md), custom slash commands, and structured context handoffs across sessions, agents and engineering roles.
- Live Course
- 5 modules
-
Running headless coding agents inside GitHub Actions and GitLab CI pipelines for automated CI failure repair, issue-to-draft-PR workflows, dependency upgrades and strict security/cost guardrails.
- Live Course
- 5 modules
-
Practical training on using IDE coding assistants such as GitHub Copilot, Cursor and Windsurf effectively through context pinning, multi-file editing, test-backed verification and secure coding habits.
- Live Course
- 5 modules
Get in touch
