Course Outline
From autocomplete to agents: understanding agent failures
• Anatomy of a coding agent: model, harness, tool surface, context, permissions
• Positioning different tools: Claude Code, GitHub Copilot, Cursor, Codex CLI, Gemini CLI
• A taxonomy of failure scenarios: incorrect context, wrong tools, lack of feedback, unbounded autonomy
Demonstration: Comparing successful and unsuccessful executions of the same task side-by-side.
Context engineering
• Managing the context window as a limited resource: determining what earns a place within it
• AGENTS.md, CLAUDE.md, .cursor/rules, copilot-instructions.md — one core concept applied across various filenames to establish a single source of truth.
• Establishing conventions, build and test commands, and defining architectural boundaries
• Retrieval versus explicit context; task decomposition and the use of sub-agents
Lab: Author repository context for an unfamiliar Python service, re-run a failing task, and compare the outputs.
Reusable workflows and Agent Skills
• Selecting the appropriate abstraction: instruction file, skill, custom command, or plain script
• Anatomy of a skill: triggering mechanisms, instructions, bundled scripts, and progressive disclosure
• Ensuring portability across tools and identifying where vendor lock-in begins
• Versioning, review processes, and distribution across a team; recognizing common anti-patterns
Lab: Build and test a reusable workflow that enforces a specific coding standard.
MCP: connecting agents to real systems
• Architecture overview: clients, servers, tools, resources, prompts; stdio and HTTP transports
• Identifying valuable servers: Git hosting platforms, issue trackers, databases, browsers, internal APIs
• Determining when a CLI or script is preferable to an MCP server
• Maintaining tool-surface hygiene: understanding why having fewer tools often leads to greater reliability
Lab: Connect MCP servers and process a ticket end-to-end—from issue creation, branching, patching, testing, to the pull request.
Feedback loops and evaluation
• Utilizing tests, type checkers, and linters as the agent's ground truth; employing test-first approaches as a control mechanism
• Leveraging CI as an outer loop and maintaining review discipline for agent-authored diffs
• Constructing golden-task evaluation sets: defining what to measure and how to detect regressions
• Treating cost and latency as primary metrics
Lab: Create a small evaluation set and score two different agent configurations against it.
Security and guardrails
• Vulnerabilities from prompt injection via issues, pull requests, READMEs, dependencies, and fetched web pages
• Implementing permission models: allowlists, approval workflows, read-only tools, and network egress controls
• Practices for secret hygiene and sandboxing: using containers, ephemeral credentials, and limiting blast radius
• Addressing 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 such incidents.
Rolling this out to a team
• A phased adoption path; deciding what to standardize versus what to leave to individual discretion
• Identifying metrics that indicate genuine value and recognizing those that do not.
Requirements
• Practical knowledge of Python, Git, and the command line
• Prior exposure to an AI coding assistant
• NobleProg will provision Dadesktop VMs for participants, pre-configured with Docker, VS Code, and Python 3.11 or later.
• A working AI coding assistant of your choice: Claude Code, GitHub Copilot, Cursor, Codex CLI, or Gemini CLI. The labs are tool-agnostic, and instructions are provided for each platform.
Audience
• Software engineers, tech leads, and architects who use AI coding assistants but struggle to achieve reliable results.
• Platform and developer-experience engineers responsible for deploying AI tooling across teams.
• Engineering managers tasked with 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
