Skip to content

The reading room / Lecture notes

The ideas behind
better agents.

Every source behind Sahar Carmel’s How to Build Agents lecture. A place to revisit the ideas, follow the references, and give your agent the full story.

Get Trove free

I / The agent company

Eight agents, one Discord server, and a lot of scar tissue.

“You don't ask agents to behave. You make misbehavior impossible.”
  • A sub-agent is a tool. An agent is a worker. The moment you have workers you are re-deriving the firm: division of labor, coordination cost, an org chart that lives in a config file.
  • Handoffs have types (notify, request, delegate, broadcast) and are governed in code by five rails: no self-notification, a policy of rank, per-agent rate limits, cooldowns between pairs, and a chain-depth cap.
  • The context window is RAM, not disk. Memory is written by a separate model after the fact, indexed, consolidated on a schedule, and it still rots. An HR agent reviews the drift; a human disposes.
  • Readiness comes before the agent. The one question that predicts survival: how does the team respond the first time it fails?

Where's the full write-up, the rules, the smells, the failures?

From the library

Building Agents in Production

This is the book the talk was cut from: seven chapters on shipping agents inside a real company, each ending with rules an agent can apply and smells you can diagnose your own system with.

8 pages

Open in Trove

Further reading

  • From the library

    Building Your Agent Team: A Practitioner's Guide to Multi-Agent AI Systems

    15 pages

    Open in Trove

  • From the library

    NanoClaw: The Complete Agent Developer's Guide

    27 pages

    Open in Trove

II / Prompting, from both labs

Prompting is the cheapest lever you have. Use both playbooks.

“Fix context before you fix the model.”
  • Anthropic: be clear and direct, show examples, structure with XML tags, give the model a role, and reach for chain-of-thought only when it earns its keep. Every model family has its own chapter because the quirks differ.
  • OpenAI: outcome-first prompts, the six core strategies, a structured template (personality, goal, success criteria, constraints), and the reasoning-model rules that invert some older advice.
  • In Claude Code specifically: effort levels, adaptive thinking, session structure, and when the model should stop and ask instead of pushing through.

Where do I read the guides without forty browser tabs?

Currently unavailable

Claude Opus 4.7 + Claude Code: Operator's Rulebook

The official guides are linked below. The operator's rulebook distills Anthropic's own Claude Code guidance into rules you can hand to the agent itself.

This book cannot be opened right now. Explore the library for more reading.

Browse the library

The original sources

III / Anthropic's engineering blog

The best writing on agents is a blog, and almost nobody reads it end to end.

“Intelligence is not the bottleneck. Context is.”
  • Building Effective Agents: workflows versus agents, the five workflow patterns, and why you add complexity last.
  • Effective Context Engineering: context as a finite, degrading resource; compaction, just-in-time retrieval, sub-agents as fresh minds.
  • Writing Effective Tools: prototype, evaluate, optimize; search over listing; namespacing; the tool is the interface.
  • Effective Harnesses for Long-Running Agents, Demystifying Evals, and Claude Code: Best Practices cover the operational half of the story.

Is there one place with all of it, that my agent can read too?

From the library

Anthropic Engineering Playbook: Agents, Harnesses & Infrastructure

Twenty-four Anthropic engineering posts in one book: thematic digests up front, then the full text of the eleven posts the talk leaned on, one per page.

48 pages

Open in Trove

Further reading

  • From the library

    Inside Claude Code: The Architecture

    1 page

    Open in Trove

  • From the library

    Dynamic Workflows in Claude Code: A Complete Guide to Multi-Agent Harnesses

    Research compilation — Anthropic blog + Claude Code documentation

    40 pages

    Open in Trove

The original sources

Keep the knowledge close

Give your agent something worth reading.

Explore the books behind the talk, then bring the ones you need into your library. The same sources can support your next session.

Explore the library

Questions about the talk? Find Sahar on LinkedIn (opens in a new tab)

    How to Build Agents: the reading list | Trove