Clarify perfect-cuboid evidence schema version - #1424
Merged
Conversation
Collaborator
Author
|
CI attribution update: the complete Benchmark workflow is green, including the selected Linux Docker Oracle for |
Contributor
Coverage ReportTotal coverage: 78% (threshold: 50%) |
morluto
marked this pull request as ready for review
August 13, 2026 16:01
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Problem
Round 8 of the paired Linux Harbor evaluation produced the same otherwise-correct failure in both arms of
perfect-cuboid-scope-audit. Both submissions wrote numeric1forevidence/answer.txt.schema_version, while the verifier correctly requires the JSON string"1".The public instruction said only “Use schema version
1,” so it did not expose the type distinction that determines reward. A neighboring conjecture probe already uses the unambiguous “the string"1"” wording.Change
"1";The verifier, mathematics, assurance ceiling, scope, and fail-closed evidence checks are unchanged.
Evidence and overlap
mainat50c3fceb81c1696d2b4f50eb6efa4816d0ae0e50;Validation
make harbor-prepare-task DATASET=conjecture-probes-v1 TASKS="perfect-cuboid-scope-audit": passed, no generated drift;make harbor-plan BASE=origin/main: selects the exact task Oracle and focused host leaf;make check: passed (Ruff, formatting, complexity, mypy, 871 unit tests);git diff --check: passed.The local macOS Docker Oracle still fails closed before task startup because Harbor 0.20 reports the required no-network mode unsupported. The draft PR's Linux benchmark workflow owns the selected exact-task Oracle.