Interview notes: how I answered a question about the agent-design concurrency model
Six rules for error handling in agent-design that I settled on
No theory here — only the problems we actually hit in production and how we got out.
After reading the agent-design core source I finally understand the trade-offs
No theory here — only the problems we actually hit in production and how we got out.
agent-design logging: what separates "readable" from "usable"
About 3000 words. Read the summary table first, then decide whether to go deeper.
Cut our agent-design build from 4 minutes to 40 seconds with these 5 changes
War story: agent-design timezone handling almost cost us a full day of billing
Open-sourced a rate-limiting middleware for agent-design with token bucket and sliding window
Conclusion first, then the full reproduction steps and raw logs. Counter-examples welcome.
Choosing agent-design: I turned 5 options into a side-by-side comparison
No theory here — only the problems we actually hit in production and how we got out.
Building an internal platform with agent-design from scratch: architecture and decision log
Happy to be convinced otherwise in the comments.
Those easily-missed type details in agent-design
About 3000 words. Read the summary table first, then decide whether to go deeper.
The edge cases the agent-design docs never spell out
I attached the relevant test cases so you can just run them.
The agent-design execution flow in one diagram (with sequence chart)
Turned it into an actionable checklist; every step says why it is done and what breaks without it.
Built a visual debugging panel for agent-design over the weekend
Conclusion first, then the full reproduction steps and raw logs. Counter-examples welcome.
This is the community home (/en/c/agent-design). Pick any item in the megamenu above and the centre column swaps to that slug’s own SFC page while both side columns stay put; switch to another community and the whole theme colour changes with it.