Summer of Context

Talk

We Tried to Detect Indirect Prompt Injection in Coding Agents. Here's What Broke

Abstract

As a community, we are delegating a significant part of our development workflow to agents, and as the use cases grow, the risk of prompt injection keeps increasing—to the point that it made it to the top of the OWASP LLM Top 10. This talk explores why detecting and preventing prompt injection is particularly hard for coding agents.

Talk

Beyond the Knowledge Cutoff: Agentic Search in Practice

Abstract

Agentic search helps AI systems retrieve current, relevant context beyond the model's static knowledge cutoff.

Talk

Your Repo Is the Memory: Durable Context for AI Coding Agents

Abstract

AI coding agents are becoming useful fast, but they still do not know the local truth of your codebase. This talk shows how to make repo-local context durable so agents produce work that fits the project and humans can still understand, review, and maintain what gets merged.