Repository navigation
docs: preserve reliability results and taxi demonstration evidence - #9
Draft
hudsonaikins wants to merge 10 commits into
Draft
hudsonaikins wants to merge 10 commits into
hudsonaikins wants to merge 10 commits into
Conversation
hudsonaikins
force-pushed
the
codex/reliability-campaign-plan
branch
from
November 13, 2026 23:51
6dcba9c to
721301f
Compare
hudsonaikins
changed the base branch from
main
to
codex/native-provider-credits
November 13, 2026 23:51
hudsonaikins
changed the base branch from
codex/native-provider-credits
to
codex/analyst-judge-context
November 14, 2026 02:04
hudsonaikins
changed the base branch from
codex/analyst-judge-context
to
codex/numeric-evidence-completion
November 14, 2026 04:02
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The owner froze BackIntel's reliability campaign. This package preserves the initial 92-response report and subsequent 46-response checkpoint, including all mandatory failures. The prepared generation-3 rerun is shelved and its $7 request withdrawn; no release acceptance is claimed.
The requested practical taxi demonstration has now run on 11,198,026 official Q1 2025 yellow-taxi records with GPT-6.1 Sol through the ChatGPT subscription. The first pass returned cannot_answer; the single repair completed its analysis and chart rendering, then failed final clean-VM export with ReadError: unexpected end of data. HTML/XLSX are not delivered. Five unchanged PNG previews were recovered from the trace and match the last successful render manifest. Three numerical claims matched independent raw-source SQL. The assessment remains needs_revision.
Two taxi attempts consumed 50 subscription model calls and 1007.818 seconds; no per-request dollar charge or paid fallback is claimed. No third attempt is planned. INT-498 tracks the observed single-report transfer defect; INT-454 tracks specific visual polish gaps.
This PR remains draft, stacked on #13, #8 and #10. Runtime source is unchanged at 9463f7cec50ccc9e, matching #13. Its 277-test implementation result is separate from live acceptance. Documentation/assessment changes passed JSON, raw-source checks, image hash comparisons and git diff --check. No merge, tag, release or deployment is performed.