AI agent playbooks

Source: https://www.instagram.com/reel/DdVargvpEQ8/?stkn=MWVxY2ljYnljdzUzcg==

AI generated summary of the linked source. Posted anonymously. Transcripts are published only for public links, never for uploads, documents or pasted text.

Your library

The AI link is plain markdown. Paste it into any assistant and it can read the whole public entry.

Summary

Cooper Simson argues that building multiple AI agents is the wrong approach, claiming Anthropic engineers dropped that method months ago. Instead, he pitches a single agent guided by "playbooks" for each task, built through interviewing yourself, storing approved outputs in a toolbox, and defining a proof checklist for what "done" looks like.

Core ideas

One context card plus 7 ideas

Multi-agent setups fail: Building a separate agent for every job and wiring them together is why most AI agent projects stall out or feel unreliable.
One agent, many playbooks: The alternative approach uses a single agent paired with distinct playbooks for each job instead of spinning up new agents.
AI delegation interview: Get Claude or ChatGPT to interview you about your job in enough depth that it could theoretically run the job without you.
Turn interviews into playbooks: That interview content gets converted into a structured playbook using a specific prompt.
The toolbox system: Every file, approved script, template, or document the agent produces gets saved and linked back to the playbook as a rule, so the agent never rebuilds the same thing from scratch.
The proof checklist: A list defining what "finished" actually looks like gets wired into the playbook so the agent checks its own work before returning it to you.
Fix the system, not the output: When something goes wrong, you update the playbook the task runs on instead of patching the answer in a chat window or retraining a giant model.

Quotes

“I don't think you should build AI agents anymore.”
@cooper.simson

“agents still feel stupid because the smartest people in AI are actually doing the complete opposite by only using one agent and then playbooks for every single job.”
@cooper.simson

“You're fixing the playbook that the task runs on. That's the whole loop.”
@cooper.simson

Resources

Links to the tools, products, repos or reading this entry mentions. Found by a web search of what the entry names, so you can go and use them.

Anyone can run this once. After that the links are saved on this entry and shown to everyone.

Remix this knowledge

Turn this entry into something you can post. Pick a format and get a fresh, original rewrite.

Remixes are AI generated, original wording, and free to reuse.

Talk it through

Open a round table on this entry and argue with other people about what it means.

Start a round table

Drop your own

Paste any video, podcast or audio link and get the transcript, the core ideas and the quotes back in one pass.

Try Drop