Browser checks run out before every criterion is seen, and the chips hide it: "pass · 19 of 19" with 3 unclear, "nothing found" after 3 fixes #354
Loading…
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
What happens
Run 166 (vLLM, stack 126) built cleanly in 93 minutes. Its browser checks still ran out before the judge had seen every criterion, and two chips hid it.
browsercheck gets 40 turns (BROWSER_STEPS), and its brief says to stop at about 25 actions (BROWSER_ACTIONS).browserchecks, on every backend.unclear: blocking the removal of a person who has expenses, Reset with Cancel, and Reset with OK. Its evidence for each says the browser check budget (4 checks) was exhausted.judge_verdictcounts every criterion, so 16 passes and 3unclearread as 19 of 19.code_review_verdictcounts only findings left open. The reviewer found and fixed three minor issues:app.tsleft ids unescaped in four attributes.What we want
UNCLEARbecause it ran out of turns doesn't count toward the cap, once. The judge can then send one continuation from the state it left. A spiral like #338's, where checks finish and disagree, still stops at four.Tab Tab Tab Enter) in one call, counted as one action;Done when
browsercall.Where to look:
BROWSER_CHECKS_MAXand its refusal insrc/osf/agent/braid_guards.js(lines 63–68 and 128);BROWSER_STEPS,BROWSER_ACTIONSand thebrowseragent's brief insrc/osf/agent/server.py(373–374 and 455–475);src/osf/engine/prompts.py(about 4985–5020);judge_verdictinsrc/osf/pipeline/fixloop.py(626);code_review_verdictand_weighedinsrc/osf/pipeline/review.py(about 2320–2372).Related: #338 (the cap), #315 (the judge's panel), #287 (review chips), #254 (apply plays the app before it's done), #350 (the backend test run 166 belongs to).
Found while reviewing run 166 (braid#350's 8-bit vLLM run), 2026-11-01. Requested by the owner.
Shipped in PR #356 (merge
c00c62f1, live on productionc00c62f).browsercheck that ran to its 40 turns no longer counts toward the session's four, once per session. A sixth check is still refused, so #338's spiral still stops. The new simbrowser-check-out-of-turnscovers it: the fourth check runs out of turns, the fifth finishes, the sixth is refused. It fails without the guard change and passes with it.pass · 16 of 19 · 3 unclear, in amber.3 fixed, or e.g.1 note · 1 fixed.tests/pipeline/test_review_verdicts.py.