admin-bypass-sweep (1) a rebase probe with a third exit code - #492
Open
EdbertChan wants to merge 4 commits into
Open
admin-bypass-sweep (1) a rebase probe with a third exit code#492EdbertChan wants to merge 4 commits into
EdbertChan wants to merge 4 commits into
Conversation
…ebase probe script whose setup failure is its own exit code. Review claim: A rebase probe that could not run exits unchecked, never conflicted. Review lane: behavior Safety invariant: A setup failure can never produce the conflict exit code; the script reuses the contract already shipped at engine/hooks/gh-write-verification/verify_pr_landed_on_trunk.sh lines 4-6 (EXIT_OK=0, EXIT_FAIL=1, EXIT_UNCHECKED=3) rather than defining a new one. Effectiveness measurement: A fixture that removes the scratch directory without deregistering it makes the probe exit 3. The same fixture under a two-branch loop yields the conflict code, so the test fails before the change and passes after. Slice rationale: One executable and its tests, with no caller changes, so the exit contract can be reviewed on its own. Architectural effect: Adds one shell script under scripts/ and its test module. Goal: Add scripts/probe_branch_rebase.sh exiting 0 on a clean rebase, 1 on a real content conflict, and 3 when the probe could not run. Motivation: During a sweep of nineteen pull requests an inline loop reported three conflicts when no rebase had executed, because a stale worktree registration made setup fail into the same branch as a genuine conflict. The same two-branch shape gated a force-push and an override merge on a clean result. Alternative considerations: Folding this into the caller was rejected because the exit contract is the reviewable unit. Making the unchecked case exit 1 was rejected because it is the exact conflation being removed. Implementation details: Create scripts/probe_branch_rebase.sh taking a source ref and a base ref. Declare the three exit codes to match engine/hooks/gh-write-verification/verify_pr_landed_on_trunk.sh lines 4-6 (EXIT_OK=0, EXIT_FAIL=1, EXIT_UNCHECKED=3). Any failure while creating the scratch worktree exits 3 with the underlying error on stderr and must not attempt the rebase. Only a rebase that actually ran may yield 0 or 1. Tear down with git worktree remove --force. Add tests/test_probe_branch_rebase.py with one case per exit code. Non-goals: Does not change any caller, does not resolve conflicts, does not alter merge commands, and does not edit any skill. Layer: domain Feature state: active Files: scripts/probe_branch_rebase.sh, tests/test_probe_branch_rebase.py Change types: - scripts/probe_branch_rebase.sh: create - tests/test_probe_branch_rebase.py: create Acceptance criteria: - `python3 -m unittest discover -s tests -p 'test_probe_branch_rebase*.py' -v` exits 0. - The probe exits 3, not 1, when its scratch worktree cannot be created. - scripts/probe_branch_rebase.sh contains no `rm -rf`. Solution: Ship a rebase probe script whose setup failure is its own exit code. Review claim: A rebase probe that could not run exits unchecked, never conflicted. Review lane: behavior Safety invariant: A setup failure can never produce the conflict exit code; the script reuses the contract already shipped at engine/hooks/gh-write-verification/verify_pr_landed_on_trunk.sh lines 4-6 (EXIT_OK=0, EXIT_FAIL=1, EXIT_UNCHECKED=3) rather than defining a new one. Effectiveness measurement: A fixture that removes the scratch directory without deregistering it makes the probe exit 3. The same fixture under a two-branch loop yields the conflict code, so the test fails before the change and passes after. Slice rationale: One executable and its tests, with no caller changes, so the exit contract can be reviewed on its own. Architectural effect: Adds one shell script under scripts/ and its test module. Goal: Add scripts/probe_branch_rebase.sh exiting 0 on a clean rebase, 1 on a real content conflict, and 3 when the probe could not run. Motivation: During a sweep of nineteen pull requests an inline loop reported three conflicts when no rebase had executed, because a stale worktree registration made setup fail into the same branch as a genuine conflict. The same two-branch shape gated a force-push and an override merge on a clean result. Alternative considerations: Folding this into the caller was rejected because the exit contract is the reviewable unit. Making the unchecked case exit 1 was rejected because it is the exact conflation being removed. Implementation details: Create scripts/probe_branch_rebase.sh taking a source ref and a base ref. Declare the three exit codes to match engine/hooks/gh-write-verification/verify_pr_landed_on_trunk.sh lines 4-6 (EXIT_OK=0, EXIT_FAIL=1, EXIT_UNCHECKED=3). Any failure while creating the scratch worktree exits 3 with the underlying error on stderr and must not attempt the rebase. Only a rebase that actually ran may yield 0 or 1. Tear down with git worktree remove --force. Add tests/test_probe_branch_rebase.py with one case per exit code. Non-goals: Does not change any caller, does not resolve conflicts, does not alter merge commands, and does not edit any skill. Layer: domain Feature state: active Files: scripts/probe_branch_rebase.sh, tests/test_probe_branch_rebase.py Change types: - scripts/probe_branch_rebase.sh: create - tests/test_probe_branch_rebase.py: create Acceptance criteria: - `python3 -m unittest discover -s tests -p 'test_probe_branch_rebase*.py' -v` exits 0. - The probe exits 3, not 1, when its scratch worktree cannot be created. - scripts/probe_branch_rebase.sh contains no `rm -rf`. Invoker-Finalize-Id: 259aacb5-064f-4dbf-a36f-f3fd4445463d
…erministic proof for the three-outcome rebase probe. Review claim: The probe's three exit codes are asserted by tests. Review lane: proof Safety invariant: Proof-only; adds no product behavior. Effectiveness measurement: The suite fails if the unchecked case returns 1 instead of 3. Slice rationale: One proof slice for this step. Architectural effect: None; verification only. Goal: Prove the three-outcome rebase probe deterministically. Motivation: The defect is a wrong exit path, so the proof exercises every exit path. Alternative considerations: Manual verification was rejected as non-deterministic. Implementation details: Execute the command below as the terminal proof. Non-goals: No product edits here; proof only. Layer: app_regression Feature state: active Exit code: 0 Invoker-Finalize-Id: a4316ef4-e3eb-4b23-bf12-3d12f74bcbb7
… that no ephemeral inter-task handoff files remain. Review claim: The workflow leaves no scratch handoff artifacts behind. Review lane: cleanup Safety invariant: Read-only; never deletes files, alters the index, or commits caller work. Effectiveness measurement: The gate fails when a plans/invoker-handoff.* or lens-*.json file is still present after the leaf tasks complete. Slice rationale: One terminal hygiene gate for the workflow. Architectural effect: None; check only. Goal: Confirm no ephemeral handoff files survive the run. Motivation: Inter-task scratch files leak into diffs and read as part of the work. Alternative considerations: Deleting them automatically was rejected; the gate reports, it does not mutate. Implementation details: Run scripts/scrub-handoff-artifacts.sh without --apply. Non-goals: No deletion, no index changes, no commits. Layer: app_regression Feature state: active Exit code: 0 Invoker-Finalize-Id: a726a99a-fef6-4971-8bb1-446252907754
…a363dbbc6-5be0c2cc — Terminal check that no ephemeral inter-task handoff files remain. Review claim: The workflow leaves no scratch handoff artifacts behind. Review lane: cleanup Safety invariant: Read-only; never deletes files, alters the index, or commits caller work. Effectiveness measurement: The gate fails when a plans/invoker-handoff.* or lens-*.json file is still present after the leaf tasks complete. Slice rationale: One terminal hygiene gate for the workflow. Architectural effect: None; check only. Goal: Confirm no ephemeral handoff files survive the run. Motivation: Inter-task scratch files leak into diffs and read as part of the work. Alternative considerations: Deleting them automatically was rejected; the gate reports, it does not mutate. Implementation details: Run scripts/scrub-handoff-artifacts.sh without --apply. Non-goals: No deletion, no index changes, no commits. Layer: app_regression Feature state: active
Bugbot couldn't run - usage limit reachedBugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit. A user or team admin can review and increase usage limits in the Cursor dashboard. (requestId: serverGenReqId_3a942efe-d38a-4d27-b106-e787f69c7a39) |
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.
Summary
This change adds a small check for testing whether a branch can be rebased before a pull request is swept.
Previously, a setup problem could look like a real merge conflict and trigger the wrong follow-up action.
The check now reports clean, conflicting, and unable-to-run outcomes separately.
The sweep procedure will call this check in a later step.
Review Claim
A rebase that could not run is reported as unchecked, never as a conflict.
Review Lane
behavior
Review Unit
engine-runtime
Safety Invariant
A setup failure can never produce the conflict exit code; the check reuses the established 0, 1, and 3 outcome contract.
Slice Rationale
This slice isolates the executable check and its tests so the outcome contract can be reviewed before any caller wiring.
Non-goals
Architecture
Before
graph TD A["scratch setup failure"] --> B["conflict outcome"]After
graph TD A["scratch setup failure"] --> B["unchecked outcome: 3"] C["clean rebase"] --> D["success outcome: 0"] E["content conflict after rebase"] --> F["conflict outcome: 1"]Test Plan
Test Plan
python3 engine/skills/make-pr/scripts/preflight.py --base origin/main— preflight passed; one engine-runtime file and one neutral test file.python3 -m unittest discover -s tests -p 'test_probe_branch_rebase*.py' -v— asserts clean, conflict, and unchecked outcomes.bash scripts/scrub-handoff-artifacts.sh— terminal hygiene check passed.Revert Plan
Revert Plan
git revert <sha>.Note
Low Risk
New standalone script and tests only; no changes to merge/sweep behavior until a later caller wires it in.
Overview
Adds
scripts/probe_branch_rebase.sh, a dry-run check that rebases a source ref onto a base ref inside a temporary detached worktree and reports the outcome without touching the main checkout.Exit contract: 0 when the rebase succeeds; 1 only when the rebase runs and leaves real content conflicts (
ls-files -u); 3 (UNCHECKED) for any setup or inspection failure (invalid refs, worktree creation, ambiguous rebase failure). Setup problems cannot masquerade as conflicts.Scratch worktrees live under
rebase-probe-worktreeswith hashed names and are removed on exit.tests/test_probe_branch_rebase.pycovers clean rebase, conflicting rebase, and a stale registered-but-missing worktree path that must exit 3. No sweep or skill wiring in this slice.Reviewed by Cursor Bugbot for commit 215a1ae. Bugbot is set up for automated code reviews on this repo. Configure here.