Live from the build loop · updated Sat 2:30p Central
Shipping right now.
I run an agent pipeline that turns issues into reviewed pull requests across my products. Every row below is a real build, with what it cost and how many tries it took. What each change does stays private; that the work happened, and how, does not.
- PRs merged, last 7 days
- 567
- agent builds shown
- 28
- verified on the first try
- 40%
- median compute per build
- $0.69
- Sat 2:00pBuild loopRan the build loop's self-testverified · try 2
- Sat 1:55pBuild loopRan the build loop's self-testfailed
- Sat 1:55pBuild loopRan the build loop's self-testrevising
- Sat 2:06pRunbookFixed a mobile layout issueverified
- Sat 2:24pRunbookFixed a screen bugverified
- Sat 1:10pRunbookAdded a new viewfailed
- Sat 1:10pAllographmerged
- Sat 1:10pAllographmerged
- Sat 12:59pRunbookUpdated page copyverified
- Sat 12:12pRunbookAdded a new screenverified · try 3
- Sat 1:53pAllographmerged
- Sat 1:43pAllographmerged
- Sat 11:40aAllographmerged
- Sat 1:18pRunbookAdded a data stepverified · try 2
- Sat 12:34pAllographmerged
- Sat 11:48aRunbookAdded a screenrevising
- Sat 1:22pRunbookShipped a featurerevising
- Sat 12:54pRunbookAdded a review stepverified
- Sat 11:34aRunbookFixed a data issueverified · try 2
- Sat 11:24aAllographmerged
- Sat 11:17aRunbookAdded a new screenrevising
- Sat 11:26aRunbookAdded an AI output steprevising
- Sat 10:24aRunbookAdded a settings screenverified · try 3
- Sat 12:25pAllographmerged
- Sat 11:13aAllographmerged
- Sat 9:46aAllographmerged
- Sat 10:57aAllographmerged
- Sat 9:31aRunbookShipped a featureverified
- Sat 8:30aRunbookAdded a settings screenmerged
- Sat 9:42aRunbookAdded a data stepverified
- Sat 8:01aBuild loopRan the build loop's self-testverified · try 2
- Sat 7:54aBuild loopRan the build loop's self-testfailed
- Sat 7:54aBuild loopRan the build loop's self-testrevising
- Sat 8:04aRunbookFixed a data display issueverified · try 2
- Sat 7:58aRunbookAdded a settings screenrevising
- Sat 10:20aRunbookFixed a data bugrevising
- Sat 7:49aRunbookAdded an import stepverified · try 2
- Sat 7:49aRunbookFixed a data issuerevising
- Sat 10:11aRunbookShipped a featureverified · try 6
- Sat 7:25aRunbookShipped a featurerevising
Hover to pause. Failures and send-backs stay in the feed.
Week by week
The last 12 weeks of the loop: how much merged, and what a typical build cost.
432last week
$0.69last week
Show as a table
| Week of | PRs merged | Agent builds | First try | Median compute |
|---|---|---|---|---|
| Aug 31 | 9 | 0 | – | – |
| Sep 7 | 15 | 0 | – | – |
| Sep 14 | 8 | 0 | – | – |
| Sep 21 | 198 | 3 | 0% | $0.92 |
| Sep 28 | 379 | 211 | 19% | $0.77 |
| Oct 5 | 489 | 176 | 30% | $1.04 |
| Oct 12 | 290 | 71 | 17% | $0.62 |
| Oct 19 | 426 | 210 | 17% | $0.68 |
| Oct 26 | 304 | 195 | 24% | $0.89 |
| Nov 2 | 558 | 167 | 58% | $0.80 |
| Nov 9 | 583 | 190 | 54% | $0.63 |
| Nov 16 | 432 | 157 | 58% | $0.69 |
How the loop works
- 01
An issue goes in
I write the issue and mark it ready. That's the only way work enters the loop.
- 02
An agent builds it
Within minutes an agent picks it up in an isolated branch, writes the change and its tests, and opens a pull request.
- 03
It gets checked
CI is the gate, and a separate review pass checks the work against the issue. Work that falls short goes back for another try.
- 04
It ships, or comes back to me
Verified work ships. Anything that needs a judgment call comes back to me instead of being guessed.
What stays private
- Titles are rewritten into generic summaries from a fixed vocabulary before they are published, then checked again by this site.
- Stealth products show only that a pull request merged, and when.
- No code, issue text, customer or repository detail is ever published.