ARTICLES · 11 pieces

Field notes on the agent landscapeexplainers, not announcements

Choose a tool, run a useful trial, and judge the result. Practical guides and working notes from the Agentlist catalog.

  1. 11

    BUYING GUIDE · 2026-10-12 · 9 min

    Email for AI agents: AgentMail and four alternatives

    Give an agent its own inbox, connect an existing mailbox, or build on an email API. Five providers and a practical trial to choose between them.

  2. 08

    COMPARISON · 2026-10-11 · 8 min

    Devin vs Windsurf: which coding workflow should you use?

    The names have changed. Compare the desktop review loop with a cloud handoff, then score the work that comes back.

  3. 09

    BUYING GUIDE · 2026-10-11 · 8 min

    Browserbase vs Browser Use vs Kernel: what will your agent cost?

    Compare browser infrastructure by completed work, with current prices, a worked example, and failure cases worth testing.

  4. 10

    SECURITY · 2026-10-11 · 8 min

    Gemini Code Assist security: what to check before you migrate

    Google has ended new subscriptions. Existing teams need a clear account of data, permissions, and the transition ahead.

  5. 05

    CHOOSING · 2026-10-07 · 6 min

    AI coding agents: a shortlist for your first trial

    Start with the way you work. Four candidates, with a concrete task to try in each.

  6. 06

    WORKFLOW · 2026-10-07 · 5 min

    Coding agents: editor, terminal, or cloud?

    Choose where the work should run, how you will steer it, and what you need back for review.

  7. 07

    EVALUATION · 2026-10-07 · 6 min

    A coding-agent trial you can actually score

    Give two candidates the same small bug. Count the work you do after the answer arrives.

  8. 01

    API · 2026-09-25 · 6 min

    Every AI agent on one endpoint

    The catalog is data, not just pages — one JSON response carries every system, its spec fields, and its last change.

  9. 02

    TAXONOMY · 2026-09-25 · 7 min

    Agent router vs model router

    Both promise to route work to the right thing — one picks a model for an inference call, the other picks an agent for a task. They're different layers, and they compose.

  10. 03

    TAXONOMY · 2026-09-25 · 8 min

    Agent framework vs agent library

    The test is inversion of control: a framework calls your code and owns the loop; a library is primitives you call from a loop you own.

  11. 04

    SIGNAL · 2026-09-28 · 9 min

    The personal agent landscape, argued on X

    A year of timeline fights about OpenClaw-shaped assistants settled into a rough consensus: chat is the interface, your hardware is the moat, memory is the product — and security is the tax nobody priced.