Working With Your Data, Safely: Files, Approvals and Memory

From the creator

So far the agent can talk about markets but cannot touch anything of yours, and nothing stops it doing something it should not. This episode fixes both. We give it file access so it reads a real portfolio.csv and writes its reports back to disk, then put a place_trade tool behind an explicit human approval. Approving every single read gets old fast, so we add auto-approval rules for the safe calls and one custom rule that clears small trades on its own. We finish with durable memory in two forms, the files the agent curates itself and the facts Foundry extracts automatically, and watch both survive a restart. 📌 This event is part of a series. Learn more here: https://aka.ms/FMTA/series 🔗 Learn more by exploring the resource: https://aka.ms/AgentHarness ⌚ Chapters: 00:00 Intro and Welcome 02:37 Session Overview: Files, Approvals, and Memory 05:52 File Access and Sandboxing 09:05 Code Walkthrough: Safe File Access 17:15 Configuring Agent Harness File Access 20:25 Live Demo: File Access and Restrictions 24:45 Understanding Approval Gates 26:22 Code Walkthrough: Approval Gates for Tools 30:30 Live Demo: Approval Gates in Action 36:08 Introduction to Agent Memory 38:45 Configuring Foundry Memory 43:25 Live Demo: Foundry Memory and User Scope 46:38 Handling Contradictions and Memory Updates 53:18 Implementing a Custom Local Memory Store 57:45 Recap and Putting It All Together #MicrosoftReactor #LearnConnectBuild [eventID:27518]

Published 2026-09-11 · Watch on YouTube ↗
Availability and playback restrictions are controlled by the original channel.