Repository navigation
Score with uv on Python 3.12; document every setting the scripts use - #17
Merged
Merged
Conversation
- score.py and screening/screen.py declare Python 3.12 (PEP 723) and run.sh and run-pass.sh call them through `uv run`. macOS's python3 is 3.9, which crashed both after a run, and run.sh hid the failure with `|| true`. - README: one-task quick start, cloud extras in the install line, a table of every key and setting the scripts ask for, scoring and troubleshooting. - run.sh and run-pass.sh name the current default release and adapter 0.3.43 in their hints, and check for uv up front. - run-pass.sh: `app` was built from the caller's `bench`, not its own. - CI runs shellcheck; methodology history lists 1.8.2 before 1.8.3. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Contributor
Author
|
End-to-end result of the README quick start, run verbatim plus
|
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Makes the runner and scorer work for an outside team on a stock machine.
Bug fixed: on macOS,
python3is 3.9.score.py(str | None) andscreening/screen.py(tomllib) crash on it, andrun.shswallows the crash with|| true, so a multi-hour run ends with no score and no error. Both now carry a PEP 723 header (requires-python >= 3.12, no dependencies) and the scripts calluv run, which uv users already have.Docs: the scripts say "see README.md" for keys the README never mentioned. The README now has a one-task quick start (the old one launched all 113 tasks), the cloud extras in the install line (the default
run.shenv is e2b), a table of every key and setting, how to score/rescore, and troubleshooting by API error code.Smaller: stale
runtime/reason-v7-2026-10/>=0.3.19hints inrun.sh/run-pass.sh;run-pass.shbuilt its Modal app name from the caller'sbench(shellcheck SC2318, worked only by dynamic scope); CI runs shellcheck; methodology history order.Verification
uv run --python 3.12 python -m unittest discover tests: 22 pass;shellcheck -S warning: clean.--n-tasks 1) from a freshuv toolinstall on an arm64 Mac with Docker, in Reason Tests: see the PR comment for the result.run.shvalidation paths (missing key, badEFFORT, missing grader key) exit with their messages.🤖 Generated with Claude Code