test: gate the tool output agents actually read - #5
Merged
Merged
Conversation
The failing-suite defect was found by hand while writing the README, not by the eight synthetic fixtures, which is a statement about the fixture set: it was built from shapes that were easy to generate, not from what an agent reads all day. Close that gap for the four most common ones. A cargo build with a warning and a type error among 420 compile lines, a tsc run with one TS2322 among 400 checked files, a pytest -v run with one failure among 400 passes, and a git diff whose real hunk sits before 60 identical generated-file hunks. Each asserts that the evidence deciding the outcome survives, that the view is at most half the input, and that the original bytes still round-trip. All four pass on the current engine at 92% to 97% reduction, so this locks in behaviour rather than changing it.
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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 failing-suite defect fixed in #3 was found by hand while writing the README, not by the eight synthetic fixtures. That is a statement about the fixture set: it was built from shapes that were easy to generate, not from what an agent reads all day.
This adds the four most common ones as a gate:
cargo buildwith a warning and a type error among 420 compile linestscwith one TS2322 among 400 checked filespytest -vwith one failure among 400 passesgit diffwith the real hunk before 60 generated onesEach asserts that the evidence deciding the outcome survives, that the view is at most half the input, and that the original bytes round-trip. All four already pass, so this locks in behaviour rather than changing it.