Your agent's plan is never its best plan. It's the one it thinks you'll like. If you're vibe coding with Opus 5.5, GPT-6 Astra, or Grok Bot, steal these 2 prompts. After it gives you a plan: "Now argue the opposite. You're a senior engineer who thinks this plan is horribly wrong. What would you build instead? Don't build it yet, just write the full plan with counters to every point you believe is wrong. Then judge both honestly: which plan wins, and what from the counter should we keep?" Before it writes any code: "Rate your confidence 1 to 10 on each part of this plan's tasks, based on what you've actually verified in code, not assumed. Under an 8, tell me exactly what you'd need to know to increase confidence. Don't do anything yet, we're only mapping and investigating." Frontier models are trained to agree with you. Making them attack their own plan is the closest thing to a second opinion you'll get from the same model. Last week this confidence mapping process caught 13 gaps in a plan I almost let my agents run. Bookmark these for your next session.

Random thoughts: @NotionHQ will be the best at executing the whole multiplayer humans + agents thing - most people struggle with “personal” agents because most consumers don’t care about *doing* things, most things for people *to do* are at work - the best use case for agents by far is knowledge work, exponentially true for multiplayer and interacting with coworkers, customers, etc - Notion is useful both singleplayer and multiplayer which is important - all of the things that matter for multiplayer, Notion already has built out: permissions, security, cloud, sharing, etc. It natively has network effects - Notion is neutral: you can use any model, any agent harness (Claude code etc) or even bring your OAI subscription - most people dont want to interact with another person’s “agent”, they want to interact with the *output* (docs, apps, etc) and use their agent to consume it — all intrinsically native to Notion - there is no more important layer for multiplayer to own than context and memory, that is quite literally company docs/workspaces - I think the grok bots interface makes the most sense for knowledge work and that’s already custom agents. Notion has all of the primitives for agents to do the work and has all the abstractions (agents) for teams/people to consume and interact with the work at whatever level they choose I think the most important thing for them to nail down in the near/medium term is the overlap between personal systems and work systems (auth, enterprise procurement, integrations, etc) but that is largely a product/UX challenge which I feel that Notion is S-tier at. In other words the bridge between consumer <> prosumer <> work/teams The current product gaps with its agents that instinct/muse have (computer/browser use, persistent state, isolated VMs/sandboxes, proactive, etc) are deterministic and the Notion AI team will either ship them in like a week or it’s already in beta I also feel like if Notion can somehow nail team comms or build a layer on top of existing comms platforms (email, slack, teams, etc) it’s game over *not affiliated btw, they are just currently the furthest ahead IMO and my usage is somehow increasing every day

in just a week, we will be gathering 100 of our passionate @NotionHQ users from around the world, known as Ambassadors, at our SF headquarters for an Ambassador Summit. for two full days, we will connect with, aim to excite, and share with Ambassadors all things going on with Notion now, and Notion in the near future. planning this Summit has been a true labor of love and i can't wait to share more of the story in the coming weeks. on a personal note, to go from attendee at last year's gathering to helping lead this year's, is beyond surreal. it has pushed me to make this the best Summit yet. let's see if it is!