David Lee's Workspace · generated Sunday 2026-07-26 8:30 AM ET
Last week — post-mortem · Jul 19 - Jul 25, 2026
21,069Impressions
73Engagements
3Posts
6,881Followers
+27Follower gain
+4%Impr. WoW
Three posts, one clear winner and two that leaked reach. The Slack-thread humor post hit 1.3x and carried the week. The 'Story Time: DARIUS Lives' post opened by apologizing for posting late and delivered a flat 0.0x. Impressions held near 3k a day on the 7.28 NY momentum, but total engagement fell to 73 and the week landed 47% below your 30-day baseline. The lesson repeats: value up front wins, buried or self-referential value dies.
Best 1.3x · 890 impressions
Every security team has had this meeting. The boss says AI security is now a quarterly goal. Everybody nods. And two people immediately open a text thread titled how in the $*#! are we pulling this off.
Relatable workplace humor plus a soft point. 'Every security team has had this meeting' pulled a 1.3x and 39 engagements. The winning formula is a recognizable team moment your audience has lived, then the identity lesson underneath it.
The 7.28 NY event is still doing work in the background. Daily impressions plateaued around 3,000 all week and peaked at 4,274 on Jul 21, your strongest sustained reach in a month. Followers grew another 27 to 6,881.
What missed
The 0.0x dud ('Story Time with Dave: DARIUS Lives') opened by apologizing for posting late and buried the payoff. Apologizing for posting signals the content is not worth your own priority, and it reached exactly 1 person.
The technical 'least privilege stops being the whole answer' post landed 0.2x. Strong argument, but a Friday slot and a thesis that took too long to arrive. Your sharpest ideas need to be in the first two lines, not the fifth paragraph.
Apply this week
Lead with the value, never with an apology or a warm-up. If the best line is in paragraph five, make it the hook.
Repeat the humor-plus-lesson formula that just won. A lived team moment earns the reach that lets the identity argument land.
Ride the 7.28 event while it is live news. A dated, in-the-moment post carrying the actual argument is your highest-leverage play this week.
4-week trend
Week
Posts
Impr.
Eng.
Best
Jun 21-27
1
5,454
84
1.3x
Jun 28-Jul 04
1
1,969
53
0.9x
Jul 12-18
2
18,391
102
1.5x
Jul 19-25
3
21,069
73
1.3x
This week — drafts + pre-mortem · Jul 26 - Aug 01, 2026
3 high-signal drafts, all grounded in real events and the newsletter arc (agentic AI security, NHI, MCP). The 7.28 NY event is the peg this week. Drafts only. You approve and publish in Supergrow.
Event day, dated anchor + full payload (7.28 NY AI Security)revised
9/10
Why now: The 7.28 NY event is your single biggest peg this week and your proven 1.5x format is a dated event anchor. This time it carries the actual argument instead of just the invite, fixing the exact miss from the Jul 12-18 review.
7.28.26. Tomorrow I get on a stage in New York and skip the part where we all agree AI is exciting.
Here is the conversation I actually want to have.
Every agent you put into production is an identity. Not a feature, not a workflow. An identity that authenticates, holds permissions, and makes decisions while no human is watching.
We spent twenty years learning to govern the human version of this. Joiner, mover, leaver. Least privilege. Access reviews we complained about but ran anyway.
Then we handed a non-expiring token to something that can spin up ten more copies of itself before lunch, and called it progress.
The agents are not the risk.
The standing, over-permissioned, unattended identity behind each one is the risk. And it is the exact problem we already fumbled once with service accounts.
If you are in New York on 7.28, come find me and let's have the uncomfortable version of this conversation.
If you are not, I still want to know. When an agent makes a call you never approved, whose identity is on the hook?
Pre-mortem (90%)
Scored 9/10. Only sub-perfect mark was Content DNA (paragraphs read slightly denser than David's single-sentence cadence). Highest-value fix applied below: broke the middle beats into standalone lines for rhythm and whitespace. Hook, originality, dual CTA all rated top marks.
Relatable workplace humor + soft lesson (the format that just won)
10/10
Why now: Mirrors this week's actual winner (1.3x, 'Every security team has had this meeting'). Supergrow flagged the exact repeat play: a chaotic team moment reacting to the AI mandate, with the NHI lesson underneath. Scored a clean 10/10.
Every security team is about to have this meeting.
Someone from platform engineering says the AI agents are live in production. Great news. Everyone claps. Then the identity person asks one small question. How many of them are there.
Silence.
Nobody knows. There is no list. There is a service account that got cloned for the pilot, then cloned again for the demo, then a fourth one somebody stood up on a Friday to unblock a deadline. Each one holds a token that does not expire and access nobody has reviewed since the day it was granted.
This is not a hypothetical. This is the machine identity problem wearing a hoodie and calling itself AI. The ratio people throw around is 45 machine identities for every human, and agents are about to make that number look quaint.
The good news is you already know how to fix it, because it is the same muscle you built for humans. The bad news is you have to actually go count them first.
So be honest. Do you even know how many non-human identities are running in your environment right now?
Pre-mortem (100%)
Perfect score. Optional lift: one bridging sentence into the 45:1 stat, or a light self-aware aside, would add personality. The 45:1 figure is framed the way David has framed it before ('the ratio people throw around'), an attributed industry figure, not a hard claim.
Post-event reflection + community CTA (vulnerability format, 1.3x-1.9x range)
10/10
Why now: Your personal-vulnerability-plus-community format has hit 1.3x to 1.9x across the last two months. Posts Wed or Thu after the 7.28 event, turning the stage into a real hallway story and reframing the problem as the org chart, not the tech. Scored 10/10.
I just came off a stage in New York where I argued that we are handing production access to identities we cannot fully control, and a room full of security leaders nodded along.
Here is the part that stayed with me after.
In the hallway afterward, three different people told me some version of the same thing. We know. We already have agents in production that nobody owns. We just do not know how to bring it up without sounding like the person slowing down the roadmap.
That is the real problem. Not the technology. The org chart. The identity risk is visible to the people closest to it and invisible to the people funding it, and the gap between those two rooms is where the next breach lives.
I have spent twenty-two years in identity and I still find this hard to say out loud in the wrong meeting. So I understand the hesitation. But somebody in your building already knows the agents are ungoverned.
Are you the person who says it out loud, or are you waiting for someone braver?
Pre-mortem (100%)
Perfect score. Optional lift: make the closing CTA slightly more explicit for comment volume, or unpack 'the gap between those two rooms' with one concrete line. Hook, originality, and Content DNA all top marks.