WORKSHOP [SEPARATE TICKETS]

Join us for additional Agentic Engineering all-day workshop lead by Victor Rentea.

When? Monday 19 October 2026 at 10:00-18:00
Where? Metropolo Hotel, Kraków

Get your tickets now:

TICKETS

WORKSHOP PROGRAM

Go from prompt & context engineering to autonomous multi-agent orchestration in one day. We’ll cover self-improving skills and subagents, spec-driven development, and the security and quality guardrails that keep AI on track. Jump in now — the gap between AI-native developers and the rest is growing fast.

Context & Prompt Engineering

• Fundamentals — 5-min History, Context Window, Harness/LLM, Compaction, Dumb-Zone, Sub-Agents
• Token-Efficiency — practices & tools for reducing AI costs
• Project Memory — best way to build your CLAUDE.md (or AGENTS.md)
• Progressive Disclosure Pattern — skills, nested context, file references, tripwires

Agent Tools

• MCP — architecture, custom MCP example
• Agent Tools — amazing things you can do via MCPs and CLIs (hands-on)

Spec-Driven Development

• When to directly prompt, plan-mode, brainstorming, or SDD?
• „Tres Amigos” requirements grilling, and handling business’ vibe-coding
• Spec review — how to avoid fatigue-LGTM, skill erosion, and cognitive debt
• Develop a feature using OpenSpec

Agentic Workflows

• Multi-Agent Orchestration Patterns
• State-of-the-Art Agentic Workflow — practical take-aways
• Optimizing Code-Reviews — quality guardrails, multi-agent / multi-model review, guided human review
• Architecture Observability — drift-protected high signal/noise files

AI Security

• Prompt Injection Techniques
• Agent Supply-Chain Attacks — Skills, MCPs, library post-install scripts, RAG poisoning
• Agent Sandboxing — permissions, auto-mode, OS /sandbox, Docker, and more

Requirements

Prior experience with ChatGPT, GitHub Copilot or (ideally) a CLI tool like Claude Code or Codex.