More The AI Native Dev episodes
Published 17 Feb 2026
Duration: 1877
Agentic development uses AI agents to automate software tasks, shifting the focus to speed and intent-driven specifications while relying on context management to achieve accurate results.
Most teams think agentic dev is about writing better prompts. It's not. Guy Podjarny and Simon Maple explain why managing context, not crafting prompt...
Agentic development represents a shift in software engineering, emphasizing speed and the use of AI agents to automate tasks that are typically done by humans. This approach relies heavily on managing context, which is essential for aligning agents, maintaining institutional knowledge, and ensuring the accuracy of their outputs. Unlike traditional software development, agentic development is non-deterministic, requiring new practices around evaluation, observation, and iterative refinement of context to ensure effectiveness.
A major challenge in agentic development is the tendency of AI agents to re-implement existing code rather than reuse it, along with their potential reliance on outdated knowledge. Managing agent systems in a cost-effective manner is also a key concern. The development process involves continuous cycles of analysis, generation, evaluation, and iteration, with context being updated as systems evolve. Successful implementation depends on effectively communicating context to agents and integrating it into the broader software development lifecycle, much like onboarding and aligning human teams.
31 Mar 2026 Why Every Developer needs to know about WebMCP Now
Alternative approaches to Large Language Models are gaining traction, with examples like Apple's offline image detection model and the WebMCPa API addressing AI agent limitations through client-side execution, lightweight local models, and streamlined web interactions while navigating challenges in scalability, cost, and dynamic content.
24 Mar 2026 Stop Maintaining Your Code. Start Replacing It
Phoenix Architecture redefines software development by treating code as disposable, prioritizing enduring system specifications, modularity, AI integration, and balance between automation and human oversight to enable safe, iterative updates and future-ready, adaptable systems.
17 Mar 2026 We Scanned 3,984 Skills 1 in 7 Can Hack Your Machine
AI skills pose significant security risks, with 13.4% containing critical vulnerabilities like prompt injections and unauthorized access, driven by high privileges and obfuscated threats, requiring tools like Sneak/Snyk and complementary measures such as code reviews and supply chain monitoring.
10 Mar 2026 The Greatest Time to Build a Startup (The AI-Native Advantage)
Agentic development uses automated agents powered by Large Language Models to enhance software engineering processes, with benefits including streamlined coding, testing, and collaboration.
3 Mar 2026 Why Your Agent Needs Memory, Not Just Context
Agents contribute to knowledge management by retaining and retrieving information through a combination of embedded models, databases, and large language models.