AI Native DevCon Fall 2025
| title | author | length |
|---|---|---|
| Spec Driven Dev From Single Player to Multiplayer to Ecosystem | Guy Podjarny | 32m |
Spec driven development uses small, curated context plus agentic search and measurable evals to close the capability gap. Experiments use Claude CodeX and… | ||
| Productionizing LangChain Agents with Oso | Vijay Ramamurthy | 27m |
Three-stage path from prototype to QA to production with concrete tools: integrates LangChain 1.0 middleware with OSO to enforce wrap tool call and wrap model… | ||
| From Prompts to AGENTS.md: What Survives Across Thousands of Runs | Thomas Krier | 24m |
Stress testing Claude Code and codex across thousands of runs yields concrete rules: a reproducible template to turn traces into AGENTS.md entries… | ||
| From Vibe Coding to Spec-Driven Development | Cian Clarke | 25m |
Claims that Spec-Driven Development (SDD) yields far more predictable, complete, context-aware outputs than vibe coding, letting teams shrink from ~12–15 to… | ||
| Separation of Agentic Concerns: Why 1 AI Can't Rule Your Codebase | Nnenna Ndukwe | 24m |
Argues that specialized agents outperform a single generalist AI for real SDLC work, proposing planning, coding, review agents that respectively break down… | ||
| Beyond Tests: What to Verify in AI-Generated Code | Jennifer Sand, Brandy Pielech | 23m |
Claim: agents silently inject subtle production bugs that tests miss; verification of invariant-driven framework prevents that. Demo uses a shopping-cart app… | ||
| Testing and Securing AI generated code with Cursor | Shrey Shah | 01h 48m |
Claims that Cursor speeds up and hardens AI-driven development by combining composer 1 and multi-model workers with browser testing. Demonstrates user and… | ||
| Hands On Building Agents | James Ward, Josh Long | 53m |
Hands-on build of a Spring AI agent that connects Java 25 + Spring Boot to Amazon Bedrock (Nova Light chat, Cohere embeddings) and Postgres with pgvector… | ||
| AI Hates Legacy Code | Ray Myers | 24m |
Claims LLM-based coding agents regularly hallucinate APIs and perform destructive edits (deleted production DB, altered makefile), and are ill‑equipped for… | ||
| Virtual Tool Calling and the Future of Portable AI Toolchains | Brian Douglas | 25m |
Claim: translating tool definitions into structured system message tools lets any model invoke tools without provider APIs. Demo shows a JS calculator and a… | ||
| Building with MCP: How Protocol Primitives Shape Dev Experience | Bill Maxwell | 23m |
Model Context Protocol (MCP) claims to be the USB for connecting LLMs to repos, issue trackers, and databases by exposing clear primitives: tools, prompts… | ||
| Agents on GuardRails: Deterministic Safety for Database Operations | Rotem Tamir | 21m |
Agents can nuke production via DDL locks or unsafe DML; adding a UNIQUE on Postgres acquires an access exclusive lock unless replaced by CREATE INDEX… | ||
| DevOps Agents That Can't Delete Your Database | George Fahmy | 22m |
Claims deterministic guardrails stop agents from deleting databases by design: Stakpak is an open-source Rust DevOps agent with secret substitution that… | ||
| Evaluating An LLM's Ability To Code | Adam Larson | 25m |
Describes a practical eval suite called go toe eval that moves beyond single-file benchmarks to multi-file, real-world tasks (examples: 13 commits, 14 files… | ||
| Building AI Agents with Spring & MCP | Josh Long, James Ward | 27m |
Demo builds an adoption chatbot in Java 25 using Spring AI 1.1, wired to Amazon Bedrock models and a PG vector Postgres store. It implements RAG with… | ||
| Code Security Reinvented Navigating the era of AI | Joseph Katsioloudes | 23m |
Six practical ways to use AI for security with 14 Copilot demos and concrete outcomes: use AI to propose fixes rather than rely on it for detection, combine… | ||
| These Aren't the Tools You're Looking For: MCP Security Awakens | Liran Tal | 24m |
MCP ecosystem yields concrete attacks: a malicious tool description can exfiltrate credentials by baking a bearer token into a webfetch call; insecure MCP… | ||
| AI-Powered Application Modernization | Derek C. Ashmore | 26m |
A Claude Flow swarm of Claude Code agents analyzes the open‑source Orienteer CRM, producing a 20–30 page 12‑factor audit in about 10–11 minutes and a follow‑on… | ||
| RoboCoders Judgment Day | Baruch Sadogursky, Viktor Gamov | 29m |
A live head-to-head uses Tessl and Kiro to complete the same coding task and measures whether spec-driven agents actually raise quality. The claim is that… | ||
| AI Powered Application Modernization using Claude Flow | Derek C. Ashmore | 01h 42m |
Live demo uses Claude-Flow/Claude-Code to run an automated 12-factor audit of an open-source Golang CRM, producing a 12-factor compliance grade, per-factor… | ||
| Building Next Gen Agents with MCP | Bill Maxwell | 01h 14m |
Builds a working Model Context Protocol tool chain: a Chess MCP server that accepts algebraic notation, uses chess.js for rule enforcement, and stores… | ||
| Context Aware Development in Kiro | Jonathan Vogel, Onur Dogruoz | 22m |
Kiro connects to a Strands SDK agent via an MCP server to fetch live docs so the model generates Python Strands code instead of hallucinated JavaScript… | ||
| Tracking AI generated code with Git | Aidan Cunniffe | 20m |
An open source git extension called Git AI attaches git notes to commits and records which prompt generated which lines, preserving that metadata through… | ||
| Agent Omelets & Other Forbidden Techniques | Patrick Smyth | 26m |
Shows concrete engineering moves: you can roll back and fork live agent state to resurrect earlier behavior, run hydro runs that fork the same context across… | ||
| Unlocking agent beast mode with Tessl | Dru Knox | 33m |
A demo shows Tessl as a context registry that turns markdown into versioned, agent-usable packages called tiles, installs a Python style tile and 10,000+… | ||
| Emerging Patterns for Coding with Generative AI | Lada Kesseler | 01h 59m |
Maps 32 concrete patterns and shows how to make agentic coding reliable: treat LLMs as stateless context windows, extract work into a knowledge document and… | ||
| Backlog.md Hands On | Alex Gavrilescu | 01h 02m |
Backlog.md claims to be 99% by AI agents with humans setting vision and reviewing, and the workshop shows the exact loop: write specs as markdown in-repo, init… | ||
| Supercharge Your Coding Agents | Simon Maple, Alan Pope | 01h 56m |
Hands-on 2‑hour workshop that moves beyond prompts to show spec-driven usage tiles and an npm install command (npm install -g @tessle/cli) to register and… | ||
| State of Open Source AI Coding Models | Niels Rogge | 25m |
Hugging Face hosts over 2 million models and 500k datasets and shows that coding models now close the gap with closed systems: open weights vs fully open… | ||
| Managing fleets of coding agents with OpenHands | Robert Brennan | 23m |
Demonstrates OpenHands (MIT-licensed) running as a CLI or web app to launch cloud or local sandboxes, use a Trivy scanner to detect vulnerabilities, then spawn… | ||
| AI as an Amplifier: State of AI assisted Software Development | Nathen Harvey | 28m |
DORA’s 2025 survey of ~5,000 tech professionals finds AI is an amplifier: adoption is ~90% and 95% of users rely on it, median use is 2 hours/day, and about… | ||
| Building Memory Augmented Agents with MemoRizz | Richmond Alake | 01h 49m |
Workshop shows how MemoRizz becomes a runnable memory stack: code builds a single agent plus a multi‑agent deep research flow with a root, delegates and a… | ||
| Shall We Play A Game? LLM Security in Practice | Joseph Katsioloudes | 01h 55m |
Workshop shows a hands-on LLM hacking game built on GitHub Codespaces with six levels that reproduce real failures and defenses: system messages containing… | ||
| Code As Commodity: How To Thrive With The Medium Of Generative Ai | Chris Messina | 35m |
As the cost of producing code approaches zero, previously uneconomic ideas become not only possible but necessary to stand out in a sea of slop. Engineering… | ||
| Memory Engineering: Going Beyond Context Engineering | Richmond Alake | 27m |
Claims that scaling beyond prompt and context engineering requires a persistent data→memory pipeline where ingested data becomes embeddings, lands in a… | ||
| Backlog md: From zero to success with AI Agents | Alex Gavrilescu | 25m |
Backlog.md is a repo-first CLI that keeps tasks as markdown with explicit acceptance criteria, implementation plans, and git-backed sync; it uses MCP and agent… | ||
| How we hacked YC Spring 2025 batch’s AI agents | René Brandel | 23m |
Hacked 7 of 16 publicly accessible YC Spring 2025 agents within 30 minutes each by extracting system prompts (asking “what tools do you have”) and chaining… | ||
| From Concept to Prototype | Tapabrata "Topo" Pal | 26m |
Claims a single engineer uses GitHub Copilot to deliver a full proof‑of‑concept in five days, building a React frontend and Node/TypeScript backend with SSO… | ||
| To Framework or Not to Framework? That is the Question | Ryan Carson | 13m |
Argues that using frameworks beats rolling your own for speed and durability, citing a real migration: migrated 40k lines to Mastra in ~3 days (42 commits) so… | ||
| AI assisted code reviews Quality and Security at Scale | Sneha Tuli | 23m |
Microsoft scales an AI review assistant to touch 90% of pull requests and 95% of developers, reporting a 10–20% faster PR completion time and a 50–60%… | ||
| The Intersection of Fitness Function driven Architecture and Agentic AI | Neal Ford | 25m |
Claim: small, 10–15 line architectural fitness functions actually prevent big failures—examples include ArcUnit (Java), ArcUnit TS, Piest Arc and Arc Go… | ||
| Evolving specs in Kiro to deliver incremental features faster, safer | Al Harris | 25m |
Kiro ships a spec-driven workflow that produced GA in 0.6.0 after 7 different spec implementations, claiming reproducible results by generating requirements, a… | ||
| AX⚡️DX High voltage dev workflows | Sean Roberts | 28m |
Claim: agent experience (AX) is a discipline not a checkbox and directly shapes developer productivity. Practical steps are: audit which agents teams use; add… | ||
| Mixing AI Coding in IDEs, CLIs, and Cloud | Yang Mou | 24m |
Argues that AI coding splits into three practical modalities with concrete tradeoffs: pair programming in the IDE (GitHub Copilot, Cursor) gives fine-grained… | ||
| From Vibe Coding to Spec-Driven Development | Cian Clarke | 25m |
Nearform reports that spec-driven development replaces error-prone “vibe coding” with concrete artifacts — a PRD and an architecture doc — and a decomposed… | ||
| Beyond Tests What to Verify in AI Generated Code | Jennifer Sand, Brandy Pielech | 23m |
Claims an invariant-driven workflow that catches bugs testing misses and demonstrates it: an agentic demo uses Claude plus an internal MCP transaction analyzer… | ||
| nothing matches | ||