Proposal review: one review step on every backend; remove the four-lens fan-out #353
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
agent.reviewfanned out overspec,scope,testandrisk;agent.review.combined(#275), which answers all four questions in one reading.one_sessionandmany_sessionsconditions. Only a one-session backend gets the single reviewer; every other backend gets four.What we want
The owner's call (2026-11-01): there should be only one review step.
agent.review.combineddoes now, whatever the parallel sessions.agent.reviewfan-out over the lenses and theone_sessionandmany_sessionsconditions.n_lenses;Done when
thoroughrun has one round-one review step, on vLLM (6 sessions) and on a one-slot llama.cpp backend alike.Where to look:
src/osf/pipeline/nodes.py:_LENS_KEYSand_COMBINED_REVIEW_NODE(~3191–3206);_THOROUGH(~3211–3226);"agent.review": {"when": "many_sessions"}in_THOROUGH_FIELDS(~3316);when:names (~228–230);src/osf/engine/conditions.py:_sessions,_one_sessionand_many_sessions(~466–497), and the registry (~997). Only these readOSF_PARALLEL_SESSIONS, whichsrc/osf/engine/scheduler.pysets in_condition_env(~2283) and in the step environment (~2750);src/osf/engine/prompts.py:review-lensprompt's single-lens and combined paths (~1175–1260,_combined_lens_part~1332);src/osf/pipeline/review.py:_merge_findings,_refutationsand then_lensesordering (~1040–1080, ~1285–1305, ~1350);MIN_LENSES_FOR_AUTOand the thin-review rule (~187, ~746–790);src/osf/engine/executors.py:_judge_combined_review(~3176, ~3279), andcombined_recordsinsrc/osf/engine/looked.py(~408);ui/src/lib/steps.ts(~142);openspec/specs/proposal-review/spec.md: line 10, "A combined reviewer covers the four lenses on a one-session backend" (~511), and approval (~86);openspec/specs/pipeline-steps/spec.md(~29);docs/guide.md, "Review is four readings, not a vote" (~1833);sim/plays/default.yaml(~117–128),sim/solo.yaml, andtests/pipeline/test_loader.py(~154).Expect: on stack 126 the four lenses took 9.7 minutes in run 166 and wrote 64,127 tokens between them. One reviewer at 45 t/s takes as long if it writes about 26,000. The first comment has run 166's numbers.
Related:
Rescoped by the owner on 2026-11-01. The first version asked for a choice between the two shapes and an A/B test. Its table of runs 160 to 166 and the outside evidence are in this issue's edit history.
Four-lens baseline: run 166, round one (stack 126, arm R, the four lenses side by side; measured from the run's transcripts)
Proposal review: let a run choose four lenses or one combined reviewer, so the two can be comparedto Proposal review: one review step on every backend; remove the four-lens fan-outDone in batch A, merged as #356 (
c00c62f1), and live on production: Braid reports commitc00c62f, which includes it. It was checked with the repo's tests and sims before the merge. A live check on vLLM is still to come: trogdor is off the network. Closed at the owner's request.