Course Outline
From autocomplete to agents: understanding agent failures
• <--[end if]>Anatomy of a coding agent: model, harness, tool surface, context, permissions
• <--[end if]>Tool placement: Claude Code, GitHub Copilot, Cursor, Codex CLI, Gemini CLI
• <--[end if]>A taxonomy of failure: incorrect context, inappropriate tools, lack of feedback, unbounded autonomy
Demonstration: Executing the same task successfully versus unsuccessfully, side by side
Context engineering
• <--[end if]>Managing the context window as a finite budget: determining what earns a place within it
• <--[end if]>AGENTS.md, CLAUDE.md, .cursor/rules, copilot-instructions.md — one concept implemented via various filenames for a single source of truth
• <--[end if]>Conventions, build and test commands, architectural boundaries
• <--[end if]>Retrieval versus explicit context; task decomposition and sub-agents
Lab: Write repository context for an unfamiliar Python service, then re-run a failing task to compare outputs
Reusable workflows and Agent Skills
• <--[end if]>Selecting the appropriate abstraction: instruction file, skill, custom command, or plain script
• <--[end if]>Anatomy of a skill: triggers, instructions, bundled scripts, progressive disclosure
• <--[end if]>Cross-tool portability and the onset of vendor lock-in
Versioning, review processes, and team distribution; common anti-patterns to avoid
Lab: Create and test a reusable workflow that enforces house coding standards
MCP: Connecting agents to real-world systems
• <--[end if]>Architecture: clients, servers, tools, resources, prompts; stdio and HTTP transports
• <--[end if]>Valuable servers: Git hosting platforms, issue trackers, databases, browsers, internal APIs
• <--[end if]>When a CLI or script is more effective than an MCP server
• <--[end if]>Tool-surface hygiene: why an abundance of tools reduces reliability
Lab: Wire up MCP servers and manage a ticket end-to-end — issue, branch, patch, tests, pull request
Feedback loops and evaluation
• <--[end if]>Tests, types, and linters as the agent’s ground truth; using test-first approaches as a control mechanism
• <--[end if]>CI as the outer loop and review discipline for agent-authored diffs
• <--[end if]>Golden-task evaluation sets: defining metrics to measure and detecting regressions
• <--[end if]>Treating cost and latency as primary metrics
Lab: Build a small evaluation set and score two agent configurations against it
Security and guardrails
• <--[end if]>Prompt injection vectors: issues, pull requests, READMEs, dependencies, and fetched pages
• <--[end if]>Permission models: allowlists, approvals, read-only tools, network egress controls
Secret hygiene and sandboxing: containers, ephemeral credentials, limiting blast radius
• <--[end if]>Supply-chain risks associated with third-party MCP servers and shared skills
Lab: Observe an agent being hijacked by a poisoned repository, then harden the setup to prevent this
Rolling this out to a team
• <--[end if]>A staged adoption path; determining what to standardize versus leaving to individual discretion
• <--[end if]>Metrics indicating genuine value versus those that do not.
Requirements
• Proficiency in Python, Git, and command-line interfaces
• <--[end if]>Prior experience with an AI coding assistant
• <--[end if]>NobleProg will provision Dadesktop VMs for participants equipped with Docker, VS Code, and Python 3.11 or later
• <--[end if]>A personal AI coding assistant of choice: Claude Code, GitHub Copilot, Cursor, Codex CLI, or Gemini CLI. Labs are tool-agnostic, with instructions provided for each option.
Audience
• <--[end if]>Software engineers, technical leads, and architects utilizing AI coding assistants but struggling to achieve reliable results
• <--[end if]>Platform and developer-experience engineers deploying AI tooling across teams
• <--[end if]>Engineering managers establishing standards, guardrails, and success metrics.
Testimonials (3)
The trainer is patient and very helpful. He knows the topic well.
CLIFFORD TABARES - Universal Leaf Philippines, Inc.
Course - Agentic AI for Business Automation: Use Cases & Integration
Good mixvof knowledge and practice
Ion Mironescu - Facultatea S.A.I.A.P.M.
Course - Agentic AI for Enterprise Applications
The mix of theory and practice and of high level and low level perspectives
