Repository navigation
Harden reviewer runtime and profile output boundaries - #6
Merged
dkropachev merged 2 commits intoOct 25, 2026
Merged
Conversation
dkropachev
changed the title
dkropachev
force-pushed
the
test/runtime-integration
branch
from
October 25, 2026 04:43
b28a189 to
3190ebe
Compare
Forward trial prompts on stdin and accept immutable local image IDs. Retain interrupted receipts and charge unfinished attempts the greater of reserved and elapsed time so cleanup cannot restore campaign limits. Remove named version probes on every exit path, propagate cleanup failures, and document recovery. Exercise output, read-only mounts, failures, timeout, SIGINT, and container removal with a fake reviewer and synthetic authentication in the Docker integration job.
Validate every frozen fixture repository and resolve distinct contained report and profile destinations before exporting private artifacts. Stage both exports in a fresh private directory so legacy temporary symlinks or hardlinks cannot redirect writes or corrupt comparison data. Cover later fixture roots, final-path aliases, destination collisions, and legacy temporary links with synthetic promotion regressions.
dkropachev
force-pushed
the
test/runtime-integration
branch
from
October 25, 2026 15:44
5f5b74b to
2a592c7
Compare
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Issue
Docker trials do not attach container stdin, so the task prompt is lost before it reaches the reviewer. Mocked subprocess coverage does not exercise this behavior or real container cleanup. Profile promotion also permits private output inside a reviewed fixture repository.
Fix
Forward stdin and add real Docker lifecycle integration with a fake reviewer and synthetic authentication. Cover output capture, read-only source, failure, timeout, interruption, and removal without inference. Accept immutable local image IDs and retain interrupted receipts explicitly.
Check all frozen fixture repository roots before writing a promoted profile, including paths reached through symlink aliases. Keep successful profile output outside the package and reviewed repositories.
Require successful native review coverage with verified bounds on all inherited host-visible input before acceptance. Historical fail-closed native evidence remains useful, but does not meet this requirement.
Decisions