Throw Away The Vibes: Context Engineering Is All You Need - DDD Melbourne 2026

We've all been there: GitHub Copilot promises to be your coding companion, but instead feels more like that overeager intern who confidently writes brilliant code for the wrong problem. As AI-assisted development tools become ubiquitous, the industry narrative promises revolutionary productivity gains. Yet many practitioners find themselves playing an exhausting game of context whack-a-mole—constantly explaining, re-explaining, and fixing what their AI "partner" confidently got wrong.

Having spent considerable time wrestling with this gap between AI promises and reality, I've discovered that the real challenge isn't prompt engineering—it's context engineering. This talk explores why "vibe coding" is fundamentally broken and introduces the systematic discipline that separates magical AI experiences from expensive disappointments. Through the constraint-context matrix and real examples like the Breadcrumb Protocol, I'll demonstrate why human expertise in scaffolding, steering, and domain understanding isn't just relevant—it's the secret ingredient that makes AI actually work. You'll leave with practical strategies for engineering context systematically and a framework for building sustainable human-AI collaboration that leverages both your skills and the machine's capabilities.

Transcript

1. Throw Away The Vibes: Context Engineering Is All You Need 21th Feb 2026 Dasith Wijesiriwardena @dasiths
2. I am here because I have a love hate relationship with distributed systems. HELLO! I AM Dasith @dasiths dasith.me
3. Context Engineering The Vibe Coding Problem Agenda Lessons & Patterns Why AI Coding Fails @dasiths
4. @dasiths “ ..Our philosophy is don't fix the code. Fix the context.. https://www.acquired.fm/episodes/how-is-ai-different-than-other-technology-waves-with-bret-taylor-and-clay-bavor https://sierra.ai
5. So, What’s the Problem With Vibe Coding? @dasiths
6. Promise vs Reality @dasiths
7. @dasiths
8. @dasiths ▪ It doesn’t work, I don’t know why. ▪ It works but I don’t know why. ▪ It works with my prompt. Growing Pains
9. @dasiths ▪ It doesn’t work, I don’t know why. ▪ It works but I don’t know why. ▪ It works with my prompt. Growing Pains
10. @dasiths ▪ It doesn’t work, I don’t know why. ▪ It works but I don’t know why. ▪ It works worked with my prompt. Growing Pains
11. @dasiths ▪ It doesn’t work, I don’t know why. ▪ It works but I don’t know why. ▪ It works worked with my prompt. Growing Pains
12. Looking at @dasiths
13. Coding Agents @dasiths “Senior implementation skills with junior design judgment” - Someone, 2025
14. Coding Agents @dasiths “Senior implementation skills with junior design judgment” - Someone, 2025
15. @dasiths https://blog.thepete.net/blog/2025/05/22/why-your-ai-coding-assistant-keeps-doing-it-wrong-and-how-to-fix-it/
16. @dasiths https://blog.thepete.net/blog/2025/05/22/why-your-ai-coding-assistant-keeps-doing-it-wrong-and-how-to-fix-it/ Granular tasks
17. Reality Check @dasiths
18. @dasiths
19. • Poisoning • Distraction • Confusion •Clash How Long Context Fails @dasiths https://www.dbreunig.com/2025/06/26/how-to-fix-your-context.html
20. Give your model the… > Right Context At the… > Right Time Let’s Fix Context @dasiths
21. Give your model the… > Right Enough Context At the… > Right Time Let’s Fix Context @dasiths https://blog.langchain.com/multi-needle-in-a-haystack/
22. Give your model the… > Right Enough Context At the… > Right Time Let’s Fix Context @dasiths https://www.youtube.com/watch?v=6_BcCthVvb8
23. Context Engineering @dasiths Foundations of
24. Context Engineering involves creating systems that provide the correct information and tools, in the proper format, exactly when needed, so that a large language model (LLM) has everything required to complete a task. The What? @dasiths https://www.philschmid.de/context-engineering
25. Context Engineering involves creating systems that provide the correct information and tools, in the proper format, exactly when needed, so that a large language model (LLM) has everything required to complete a task. The What? @dasiths https://www.philschmid.de/context-engineering
26. @dasiths
27. 1. Write Context 2. Select Context 3. Compress Context 4. Isolate Context Core Tactics @dasiths https://rlancemartin.github.io/2025/06/23/context_engineering/ https://www.dbreunig.com/2025/06/26/how-to-fix-your-context.html
28. 1. Write Context 2. Select Context 3. Compress Context 4. Isolate Context @dasiths https://rlancemartin.github.io/2025/06/23/context_engineering/ https://www.dbreunig.com/2025/06/26/how-to-fix-your-context.html Core Tactics
29. 1. Write Context 2. Select Context 3. Compress Context 4. Isolate Context @dasiths https://rlancemartin.github.io/2025/06/23/context_engineering/ https://www.dbreunig.com/2025/06/26/how-to-fix-your-context.html Core Tactics
30. 1. Write Context 2. Select Context 3. Compress Context 4. Isolate Context @dasiths https://rlancemartin.github.io/2025/06/23/context_engineering/ https://www.dbreunig.com/2025/06/26/how-to-fix-your-context.html Core Tactics
31. 1. Write Context 2. Select Context 3. Compress Context 4. Isolate Context @dasiths https://rlancemartin.github.io/2025/06/23/context_engineering/ https://www.dbreunig.com/2025/06/26/how-to-fix-your-context.html Core Tactics
32. @dasiths
33. ▪ Inconsistent Implementation – “The overeager intern” ▪ Knowledge Silos – “When the why is missing” ▪ Progress Fragmentation – “Worked with my prompt” @dasiths https://dasith.me/2025/04/02/vibe-coding-breadcrumbs/ The Misalignment Problem
34. The cost of this misalignment grows as development continues. ▪ Code reviews become more difficult ▪ Onboarding new team members takes longer ▪ The AI becomes less effective as a collaborator over time. @dasiths https://dasith.me/2025/04/02/vibe-coding-breadcrumbs/ The Misalignment Problem
35. The solution lies in creating an external, persistent shared context that both humans and AI can access and update. It’s a structured workflow consisting of… - Structured Planning & Task Management - Centralized & Accessible Knowledge Context - Living Documentation & Shared Understanding @dasiths https://dasith.me/2025/04/02/vibe-coding-breadcrumbs/ Solution: Breadcrumb Protocol
36. https://github.com/dasiths/VibeCodingBreadcrumbDemo
37. @dasiths Solution: Breadcrumb Protocol
38. @dasiths Research Embrace Plan Implement
39. AI coding assistants are brilliant at simple tasks and break everything they touch on complex ones. The root cause: AI can't tell the difference between investigating and implementing. When you ask for code, it writes code. It doesn't stop to verify that patterns match your existing modules or that the APIs it's calling actually exist. You: "Build me a Terraform module for Azure IoT" AI: immediately generates 2000 lines of code Reality: Missing dependencies, wrong variable names, outdated patterns, breaks existing infrastructure @dasiths Why Use RPI https://github.com/microsoft/hve-core/ Failure Mode
40. The solution isn't teaching AI to be smarter. It's preventing AI from doing certain things at certain times. This example RPI flow works by separating AI work into four distinct phases, each handled by a specialized agent: • Task Researcher: investigates your codebase and external sources, producing verified findings with citations • Task Planner: transforms research into actionable implementation plans with clear success criteria • Task Implementor: executes plans methodically, following established patterns discovered during research • Task Reviewer: validates implementation against specifications, checks compliance, and identifies follow-up work @dasiths Using RPI https://github.com/microsoft/hve-core/
41. • AVOID THE "DUMB ZONE. That’s the last ~60% of a context window. Once the model is in it, it gets stupid. Stop arguing with it. NUKE the chat and start over with a clean context. • SUB-AGENTS ARE FOR CONTEXT, NOT ROLE-PLAY. They aren't your "QA agent." Their only job is to go read 10 files in a separate context and return a one- sentence summary so your main window stays clean. • RESEARCH, PLAN, IMPLEMENT. This is the ONLY workflow. Research the ground truth of the code. Plan the exact changes. Then let the model implement a plan so tight it can't screw it up. • AI IS AN AMPLIFIER. Feed it a bad plan (or no plan) and you get a mountain of confident, well-formatted, and UTTERLY wrong code. Don't outsource the thinking. • REVIEW THE PLAN, NOT THE PR. If your team is shipping 2x faster, you can't read every line anymore. Mental alignment comes from debating the plan, not the final wall of green text. • GET YOUR REPS. Stop chasing the "best" AI tool. It's a waste of time. Pick one, learn its failure modes, and get reps. @dasiths Lessons From HumanLayer https://www.youtube.com/watch?v=rmvDxxNubIg
42. @dasiths “Harness Engineering” https://martinfowler.com/articles/exploring-gen-ai/harness-engineering.html
43. @dasiths https://blog.thepete.net/blog/2025/05/22/why-your-ai-coding-assistant-keeps-doing-it-wrong-and-how-to-fix-it/
44. @dasiths https://blog.thepete.net/blog/2025/05/22/why-your-ai-coding-assistant-keeps-doing-it-wrong-and-how-to-fix-it/
45. @dasiths Human
46. @dasiths “As we take charge of bridging the gap, the machine claims the praise.” > We Often Overestimate Machines and Undervalue Ourselves Using large language models (LLMs) for programming requires many supporting skills. However, these skills are not clearly defined and depend a lot on the situation, so right now, the best way to improve is through lots of practice and learning from mistakes. https://ferd.ca/the-gap-through-which-we-praise-the-machine.html
47. ▪ Start with one context tactic and measure results ▪ Build systematic practices around context management ▪ Recognize context engineering as core technical skill @dasiths Practical Takeaways
48. ▪ Start with one context tactic and measure results ▪ Build systematic practices around context management ▪ Recognize context engineering as core technical skill @dasiths Practical Takeaways
49. Context engineering is a systematic discipline Skilled practitioners will define the future of AI-assisted development @dasiths
50. Title Description URL [1] Pete Hodgson's Constraint-Context Matrix Primary problem framework https://blog.thepete.net/blog/2025/05/22/why-your-ai-coding-assistant-keeps-doing-it-wrong-and-how- to-fix-it/ [2] Drew Breunig's "How to Fix Your Context" Six practical tactics for context management https://www.dbreunig.com/2025/06/26/how-to-fix-your-context.html [3] Lance Martin's Context Engineering Four-strategy foundation https://rlancemartin.github.io/2025/06/23/context_engineering/ [4] Philipp Schmid's Context Engineering Context vs. prompt engineering distinction https://www.philschmid.de/context-engineering [5] Anthropic Research Context offloading (54% improvement) https://www.anthropic.com/engineering/claude-think-tool [6] "Less is More" Paper Smart tool selection (44% improvement) https://arxiv.org/abs/2411.15399 [7] Fred Hebert's Human-Machine Gap Operator skills importance https://ferd.ca/the-gap-through-which-we-praise-the-machine.html [8] Microsoft Azure CTO Warning Industry reality check on vibe coding hype https://www.geekwire.com/2025/reality-check-microsoft-azure-cto-pushes-back-on-ai-vibe-coding-hype- sees-upper-limit-long-term/ [9] Domain Understanding Requirement Context engineering needs domain expertise https://medium.com/@springrod/context-engineering-needs-domain-understanding-b4387e8e4bf8 [10] Breadcrumb Protocol Foundation workflow https://dasith.me/2025/04/02/vibe-coding-breadcrumbs/ [11] Manus Production Context Engineering Real-world lessons from building AI agents at scale https://manus.im/blog/Context-Engineering-for-AI-Agents-Lessons-from-Building-Manus References
51. Any questions? THANKS! @dasiths dasith.me https://www.nationalgeographic.com/travel/destinations/asia/sri-lanka/
53. Presentation template designed by powerpointify.com Special thanks to all people who made and shared these awesome resources for free: CREDITS Photographs by unsplash.com Free Fonts used: https://www.fontsquirrel.com/fonts/oswald @dasiths