Topics
The archive by subject, ordered by what I work on now rather than by how many posts each one holds.
If you came from the paper, the two terms it introduces have pages of their own: pre-action authorization and the Open Agent Passport.
- authorization5
Deciding what an autonomous agent may do, and enforcing it before the tool call rather than reviewing the log after.
Latest: 45% of AI Agents Share the Same API Key. Here's Why That's a Fire Waiting to Happen.
- agents20
Building and running autonomous agents: identity, delegation, lifecycle, and the failures that follow from getting those wrong.
Latest: 45% of AI Agents Share the Same API Key. Here's Why That's a Fire Waiting to Happen.
- security32
Incidents, CVEs and red-team results from agent systems, with the root cause each one shares.
Latest: 45% of AI Agents Share the Same API Key. Here's Why That's a Fire Waiting to Happen.
- mcp3
The Model Context Protocol, its server ecosystem, and the authorization gap it leaves open.
Latest: Microsoft Shipped MCP Governance. Your Agent Needs One Too.
- ai19
Where machine learning meets the systems it is dropped into, and what breaks at the seam.
- engineering9
Working notes on shipping software, mostly about the parts that resist being tidy.
Latest: I Opened an AI Delegation Spec PR at 3 AM. Here's What It Solves.
- story23
Personal writing: immigration, money, hackathons, and the years before any of the above.
Latest: Agent Registries & Kill Switches: ship trust in milliseconds