The Agent Security Checklist I Use Before Connecting Tools
Before an AI agent gets tools, files, or APIs, decide what it can read, write, call, log, and roll back.
Read postConnect your tools, let agents do the work
From the blog
Practical writing on link gmail, slack, github, notion, and more so your ai can read, draft, schedule, and act inside the tools you already use.
Agent Identity Is the Missing Security Layer for AI Workflows
Agents need verifiable identity, scoped capabilities, revocation, and audit trails before they can safely act across your tools.
Read postAgent Architecture: Multi-Step AI Workflows That Survive Production
Loop patterns, state management, and error recovery for multi-step agents that move work between tools without falling over at scale.
Read postAgent Replays with TraceTrail: Loom for Agent Runs
Agent runs are opaque. Turn a run into a stepped timeline of messages, tool calls, and tokens you can audit and share.
Read post