Repository navigation
feat(evals): unscripted multi-plugin routing runner; close the host's stdin - #6
Merged
Merged
Conversation
… stdin - scripts/run_natural_evals.py sends a case prompt as written with every given plugin loaded and read-only tools, and reports which skills the host loaded on its own over a number of runs - run_host_evals.execute no longer lets the host inherit an open stdin: Codex then waited for more input and the run timed out without one event Closes #5. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Closes #5.
Summary
scripts/run_natural_evals.py: sends a case prompt exactly as written, with every plugin given by--plugin-rootand--with-pluginloaded and read-only tools, and reports which skills the host loaded on its own over--runsrepetitions. Cases live inevals/natural/cases/, expectations inevals/natural/expected.json; expectations never enter prompts. A load is a successful skill call or a complete read of aSKILL.md, not a mention.run_host_evals.executepassesstdin=DEVNULL. A host that inherited an open stdin printed "Reading additional input from stdin..." and never started the turn; the run then timed out without a single event. This is the cause of the Codex timeouts seen earlier today.Verification
python3 -m unittest discover -s tests -t .: 181 tests pass (10 new).python3 scripts/validate-plugin: exit 0.🤖 Generated with Claude Code