Skip to main content

My reading

Pick up a thread, follow a route, or return to an older note.

Explore notes

Reading history stays in this browser only. It does not sync across devices and is lost when browser data is cleared.

Loading local reading history…

Pick up where you left off

    Agent architecture: a starting route

    One route through this topic: execution, tool boundaries, context, memory and skills. It is not a site-wide curriculum; an opened note is not a mastered skill.

    1. Harness: Controlling an Agent’s Execution
    2. Bounded Agent Loops
    3. Tool Contracts for AI Agents
    4. Context Engineering for Agents
    5. Agent Memory: Retention, Revision and Expiry
    6. Agent Skills

    Related reading

    Connections to your three most recently opened notes. Previously opened suggestions appear oldest first; unvisited ones are invitations to explore.