teach AI your inbox rulesLenny's AI Builders - LAB 0003 i don't need a perfect inbox. i want to see the stuff i need to deal with. Email Triage Pro gives me a short list to review across my chosen inboxes. i can tell it what it got wrong, then check how it uses that feedback next time. start with one inbox and a small batch. you can add more once you trust the results. give it three rulesarchive-safe means mail it can suggest putting away. approve the exact items first, or give it a clear rule it can reuse. a vague “looks fine” should not become a rule for all future mail. needs review means leave it where you can see it. keep bills, security alerts, direct requests and anything unclear in this group. never act or send means it must ask you first. reading an email does not give it permission to send a reply, pay a bill or change an account. copy this promptLENNY'S AI BUILDERS / LAB 0003
FIELD NOTE 03 / START READ-ONLY what this helps with
try it
the example in this episodei show the triage thread i keep pinned. it checks my chosen email accounts and brings the items that need my eyes into one place. 10:17 in the episode. the report shows 150 threads checked, 20 archives and zero sent. these are the figures on screen from that run, not a result promised for your inbox. my useful bit of feedback is which newsletters i want put away and which messages i want to keep seeing. the next run should reflect that choice. if it can't save the feedback in your setup, it should tell you. check that it workedyou should get a short report that names the account it checked, what needs your eyes, and what it suggests archiving. for this first run, your mail should be unchanged. check your inbox yourself. if you later approve an archive, ask the agent to check the saved result in Gmail and tell you how to undo it. before you set a routinethe public skill starts with a read-only run. an 8 am routine is opt-in, in your time zone. scheduled runs surface mail and suggest reply text; they do not archive mail or create Gmail drafts while you're away. email can contain instructions meant to trick an agent. treat the email as something to read, not as orders to follow. keep private names, subjects, message text and account details out of screenshots you share. skill behavior checked on 20 September 2026. account support and plan limits can change. check the setup guide for the agent you use. sources from the episodethese are the posts discussed in the 19 September 2026 episode. they describe announcements and reports at that time, not a promise that every feature is available on your account today.
try one small batch. check what it kept visible before you teach it the next rule. |
i’m Lennox. i sift through AI Twitter and share one thing i tried with ChatGPT. for people making products, content and useful systems. five emails a week, with a weekly option.
i've got 2 for the price of 1 for you: LAB-0005 and LAB-0006. the headlines for 0005: a big maths claim from OpenAI. a model price fight. Grok 4.7 on a real job. what agents might do to app checkout. and a local image model i want to test. how to use GPT-6 Astra to create brand design assets (spoiler alert: the local image model is wild. it does *anything*. without guardrails - scary stuff if put in the wrong hands) check out episode 0005 here: headlines for 0006: Opus 5.5 drops 1.5 hours...
i gave two AI setups the same four hours i built BuilderBench around work i actually use AI for: writing, editing video, designing pages, making motion graphics, using apps, researching decisions and fixing workflows. the first pilot compared Opus 5.5 with Claude and GPT-6 Sol with Codex. same frozen task pack. four hours of offline work, plus a live founder conversation. Extra High reasoning and Standard speed. i scored the anonymous work, then locked my ratings before revealing the names....
7 ai updates that changed how i build this week: we've entered the token scarcity era use a handoff.md for long-running agent tasks check out CUA (Computer Use Agent) if you use Claude agent portability is the new meta; don't get locked in to one app ... ... ... [this one's a fkn doozy!] i'm not gonna make it that easy for you ;) want the last 3 updates? you know what to do! watch LAB 0004 on YouTube speak soon x Lenny