Repository navigation
test: keep the command-cron pin live on sandbox-less POSIX hosts - #16096
Conversation
Address Opus review of #16083: probe the command shell after the recorder patches wrap_argv (unpatched, the probe raises on a POSIX host with no sandbox backend and wrongly skips), and re-record thread_replies_text in the frozen build_message signature so the snapshot describes main exactly.
Design Review (Fable 5) — ✅ PASSDesign-level review of Design-Verdict: PASS Verify the Linux/macOS shards of this round ran (not skipped) [DESIGN-REVIEWED] 19b9ece Description read[DESCRIPTION-READ] 16d4310a7e30624c843fa5588f089f3b9457ab7ef169e9fd60d340b1d8bc42fb That sha256 names everything this verdict read: the PR description, plus the 1 evidence file(s) the workflow collected from it and handed this review. It covers that evidence because the media strip replaces every attachment URL with the same placeholder, so an edit swapping one attachment for another would leave the description's captured text identical while what this review judged changed. Recompute it with |
GPT 5.6 Review — ✅ no blocking findingsGPT 5.6 completed its review of This comment is updated in place on each push. Review detailsNo findings. False positive or not applicable? A repository writer can comment: |
Opus 5 Review — ✅ no blocking findingsReviewed Verdict parsed from the review's SHA-scoped output markers for commit False positive or not applicable? A repository writer can comment: |
Problem / Motivation
Goal: the two tests fixed in #16083 keep guarding what they guard on every host that can run them.
Opus reviewed #16083 after it merged and found two gaps:
wrap_argv. On a POSIX host with/bin/shbut no OS sandbox backend, the unpatched probe raises insidewrap_argv, gets cached as "not a POSIX shell", and the test skips -- on exactly the host it was written for, with a false reason.build_messagesignature does not listthread_replies_text, so the snapshot no longer describes main exactly (the comparison tolerates it, but the table is wrong).Why it matters
Gap 1 silently stands down the command-cron half of the apps-tree-mask pin on sandbox-less POSIX hosts. Gap 2 leaves a pin table that misdescribes the surface it pins.
Not a goal
What changed (motivation → approach → change)
test/test_app_script_cron_sibling_import.py: the probe moves into the test body, after_record_spawnspatcheswrap_argv, with_POSIX_STRICT_CACHEcleared first so an earlier unpatched verdict cannot leak in. The probe's own spawns are cleared from the recorder before the command runs. A shell-less host (Windows) still skips; a sandbox-less POSIX host now runs the case. The real-sandbox command case also gates on the (unpatched) probe, so a host with a backend but no passing shell skips instead of failing on the refusal.test/test_context_composition_contract.py:thread_replies_text: 'str | None' = Noneis recorded in its declared position in_BUILDER_MEMBERS["build_message"].docs/system-specs/common/testing-conventions.md: one sentence -- run a gating probe under the same patches the test body uses.Backwards compatibility
Compatible: test and doc changes only.
Tests
On Windows: both files
125 passed, 3 skipped(the command case skips, as before);--noconfteston the cron file passes; flake8, black and the comment-history gate are clean. The sandbox-less POSIX path is proved by the Linux/macOS shards of this round, which must run (not skip)TestACommandCronRunningBundleCode::test_the_call_site_leaves_the_bundle_readable.Manual verification
N/A — unit coverage sufficient; nothing user-visible changes.
Related Issues
Follow-up to #16083.
Checklist