AI Agent Incident Case Studies
To close: Decide whether a case studies index page is still wanted. · Write the Outcome. · Move the cycle to HISTORY.md.
last activity May 27
One disaster story could look like bad luck. Three, each going wrong in a different way, show the problem is common.
Three case study pages are live: PocketOS, Replit and SaaStr, and AWS Kiro. The planned index page listing all three was never built.
What changed
The change that matters most, in plain words. Open any one to see it.
- 1
Three case study pages are live: PocketOS, Replit and SaaStr, and AWS Kiro.
Three stories, three failures · came from Mark Parson
- Every case study follows the same order: setup, timeline, analysis, impact, prevention and sources. In words
- The problem page shows all three as teasers that link to the full stories. In words
How we got here
Each change, traced back to the conversation it came from. A solid mark means the record names the source; a dashed one was matched by the draft and waits for review.
From the record: no recorded conversation behind this one.
No ruling needed: the work itself is the change.
Three case study pages are live: PocketOS, Replit and SaaStr, and AWS Kiro.
Thanks to
People outside Qpoint first. Every thank-you points at something they said or made.
Built the case study library
Committed the three pages and their cross-links. The record names nobody, so this rests on git authorship.
All images
Every picture from this cycle. Click any one to step through them all.
The pairs
Everything captured · 0
The full record 0 decisions · 0 key-change entries · 483 words · 0 captures
The record stays as written, for agents and for audit: Key Changes, every decision, the artifacts and the visual record.
Open the record for c4 →