Fixed: billed calls before the service recorded reviews are costs too, so the chart keeps them; the comment, the README, the PR document and the Werkbaum now say "the oldest period with reviews or…
Fixed: the Non-Goal now names the reason: the excerpt selection uses GPT-6 Luna, whose calls did not hang in the measurement.
Fixed: judging the findings and combining risks and findings are now steps 8 and 9 of their own, after the coverage check, in the order of the diagram.
Fixed: judging the findings and combining risks and findings are now steps 8 and 9 of their own, after the coverage check, in the order of the diagram.
Fixed: the sentence now names the number of pull requests, as the Werkbaum node does.
A question without "?" never passed: the pattern always needs the question mark. But a new test showed a real gap: after the "?" it allowed several sentences. The pattern now allows exactly one…
Fixed: Additional Changes now also name the planned node that keeps MiniMax M3 within its timeout.
Fixed: the class is before the __main__ block now. tests/test_doccheck.py and tests/test_risk_with_issue.py had the same mistake from PR#50 and PR#53 and are fixed too; a direct run of each…
Fixed: the check is strict again. A question must end with "?", or with "?" and exactly one sentence that delimits it, as the measured question of data_correction does; a sentence without a…
Fixed: the node is [^] now, and the legend says that [^] means live on mih09 or in use, which covers a plan in the repository.
No wait time was hidden: the entry was dropped only when no PR document was judged, i.e. without a Jev call; the other phases are entered only when they call a service. To make that explicit, the…
Fixed: the Problem section now says that this PR implements the rule for the same file, and the Non-Goals exclude RED for the same category in different files.