* test(#2596): failing-first suite for worktree-wave scope conformance Binds the advisory diff-vs-declared-scope check to behavior before it exists: the pure coverage predicate, the SUMMARY-artifact exemption and its parity with the rescue walker, the gauntlet integration (never flips ok, degrades on a git failure, survives a later block), the manifest normalizer's files_modified handling, and the --files negative-input matrix on record-agent/create. Refs #2596 * enhance(#2596): warn when a wave branch commits outside its declared scope The worktree-wave merge gauntlet validated branch, base, deletions, SUMMARY rescue and a clean worktree, but never compared a plan branch's actual committed diff against the files_modified the plan declared — so an executor that committed outside its brief merged into shared phase state silently. Adds an advisory scope-conformance check: when the manifest entry carries a declared scope, the gauntlet diffs HEAD...<branch> and appends one structured warning per path outside it. It never flips ok and never blocks the merge; promotion to a hard gate is a separate, disclosed change. With no declared scope no git subprocess is spent at all. Refs #2596 * docs(#2596): document the advisory worktree-wave scope-conformance check Records the optional --files flag on worktree record-agent/create, the advisory warnings channel cleanup-wave now emits, and its two deliberate noise limits (SUMMARY-artifact exemption, literal-prefix glob matching). Wires execute-phase to pass the plan's already-parsed PLAN_FILES. Refs #2596 * fix(#2596): close review findings on the scope-conformance advisory - share one path normalizer between the SUMMARY-artifact predicate and the scope comparison so the exemption and the check cannot drift - wire --files into the orchestrator-worktree dispatch, which created a worktree but never declared its scope, so the advisory silently did not apply on that backend; ADR-1239 requires both adapters share one check - correct the now-false blockquote claiming the check does not exist yet - add the fast-check property tests the repo requires for parser logic - add the record-agent/create parity test that Generative Fix Divergence requires for two surfaces implementing one rule Refs #2596 * fix(#2596): keep execute-phase.md under the frozen pre-phase-6 byte ceiling The one-sentence note added with the --files flag pushed execute-phase.md to 93708 bytes, past the ADR-857 PRE_PHASE6 cap of 93600 — the tightest of the three workflow size gates, and a hard cap an acknowledgment cannot clear. It failed three tests plus the differential attribution check. Condense the note to a one-line pointer (93543, 57 B of headroom); the full explanation already lives in docs/CLI-TOOLS.md and the dispatch step. The flag itself stays in the command, because the orchestrator reads this workflow at runtime and cannot pick it up from docs/. Acknowledge the remaining 143 B of growth by appending to the existing execute-phase.md fragment rather than adding a second one — the ack lint rejects two sources naming the same path. Refs #2596 * fix(#2596): make the execute-phase.md edit net-negative, not merely under the cap The size gate on this file is two assertions, not one: bytes < 93600 AND bytes <= 93400. The base is exactly 93400, so the file is at its budget and any growth trips the margin assertion — the previous fix cleared the ceiling but not that. Move the --files explanation to per-plan-worktree-gate.md, which already owns PLAN_FILES and carries no cap, and reclaim the rest from two clauses in the sentence being edited: the cleanup-wave rules phrasing, and a 'non-zero exit' the very next sentence already states. execute-phase.md ends at 93392, eight bytes below base. The flag itself stays in the command — the orchestrator reads this workflow at runtime and cannot pick it up from docs/. With no growth left, the acknowledgment is unnecessary and its byte delta was no longer true, so the shared ack fragment is restored byte-identical to base. Refs #2596 * docs(#2596): add the how-to for interpreting scope-conformance warnings The docs for this change were entirely Reference — the flag and the warning codes — with the task-oriented quadrant empty. Adds the page that answers the question an operator actually has when the advisory fires: what the two codes mean, that nothing is blocked so there is no failure to hunt for, how to tell whether the executor over-reached or the plan under-declared, and the three ways the check legitimately stays silent so an absence of warnings is not mistaken for proof of conformance. Refs #2596 * chore(#2596): backfill changeset pr number to 3264 --------- Co-authored-by: sim <sim@local>
5.7 KiB
How to interpret scope-conformance warnings
Goal: Understand the advisory warnings that the worktree.cleanup-wave gauntlet emits when a plan branch commits changes outside the scope it declared, and decide what — if anything — to do about one.
Prerequisites: GSD Core is installed and you have run /gsd-execute-phase with worktree isolation enabled (workflow.use_worktrees: true, the default), so that plan branches were merged back by a cleanup-wave.
What you will see
When /gsd-execute-phase runs plans in isolated worktrees, each plan branch is merged back into the phase branch by the worktree.cleanup-wave gauntlet. If the wave manifest recorded the plan's declared scope (files_modified, passed as --files to worktree record-agent / worktree create), the gauntlet compares the branch's actual committed diff (HEAD...<branch>) against that declared scope and reports any committed path that falls outside it.
A realistic cleanup-wave result with one such warning:
{
"ok": true,
"reason": "wave-cleanup-complete",
"warnings": [
{
"code": "scope_out_of_declared",
"branch": "plan-04-add-rate-limiter",
"path": "src/util/rate-limiter-cache.cts"
}
],
"entries": [
{
"branch": "plan-04-add-rate-limiter",
"status": "merged_removed",
"warnings": [
{
"code": "scope_out_of_declared",
"path": "src/util/rate-limiter-cache.cts"
}
]
}
]
}
ok: true and status: "merged_removed" are unchanged by the warning — the merge happened. The same warning object appears twice: once nested under the offending entry, and once aggregated into the top-level warnings array so you can scan for conformance issues across the whole wave without walking every entry.
Why this happens
/gsd-execute-phase gives each plan a declared scope up front (files_modified in the phase plan index), so that parallel plan branches can be reasoned about and merged with some confidence about what each one touched. The scope-conformance check is a post-hoc, best-effort verification of that promise: after a branch merges, the gauntlet diffs what was actually committed against what was declared, and flags any mismatch as a warning. This is issue #2596.
The two warning codes
scope_out_of_declared— emitted once per committed path that falls outside the plan's declared scope. A branch with three unexpected paths produces three of these warnings.scope_check_unavailable— emitted once per entry when the scope diff itself could not be computed (agitfailure or a timeout), rather than once per path. This means conformance is unknown for that entry, not clean — the check deliberately distinguishes "could not verify" from "verified and clean" so an unknown result is never silently read as a pass.
It does not block
Scope conformance is advisory by design. A scope_out_of_declared or scope_check_unavailable warning never changes ok, reason, the per-entry status, or the process exit code — the merge proceeds exactly as it would with zero warnings. There is no failure to fix here. Promoting scope conformance to a hard gate (one that blocks the merge) would be a separate, explicitly disclosed change — it is not what this check does today.
What to do about one
- Compare the reported
pathagainst the plan's declaredfiles_modifiedin the phase plan index. - Decide which side is wrong: either the executor committed something outside its brief (over-reach), or the plan's
files_modifiedunder-declared what the work actually needed to touch. - If the plan under-declared its scope, widen
files_modifiedin the plan so future waves report accurately. - If the executor over-reached, review that path's changes specifically — it is already merged into the phase branch, so treat the review as catching it before it travels further (into a PR, a release, or a later phase).
When nothing is reported
Absence of scope_out_of_declared / scope_check_unavailable warnings is not proof that a branch stayed in scope. The check legitimately stays silent in three cases:
- No
--fileswas recorded for the plan. Scope is unknown, so no comparison runs at all — not even agitcall is made for that entry. - Every out-of-scope path is a
.planning/**/*SUMMARY.mdartifact. These are always exempt: the executor writes them by orchestration contract, and no plan declares them as part of its scope, so flagging them would be noise rather than signal. - A declared pattern had no literal prefix (for example,
*.md). Matching is prefix-based (see below), so a pattern with no literal prefix cannot usefully bound anything — the check suppresses warnings for that entry by design, so an advisory check never cries wolf on a pattern it cannot meaningfully evaluate.
Known limits
- Glob matching is literal-prefix only.
src/**/*.tsmatches anything undersrc/, includingsrc/a/b.json— the check does not parse glob syntax past the literal prefix. - Renames are not detected specially. A rename appears in the diff as an add plus a delete. In practice this rarely reaches the scope-conformance check at all: the pre-existing deletions guard blocks any entry whose diff contains a deletion before the scope-conformance check runs.
/gsd-quickworktrees have no plan-declared scope. They are never checked, for the same reason as the "no--filesrecorded" case above.
Related
- CLI Tools reference — worktree commands — the
worktree record-agent/worktree create--filesreference - Execute a phase
- Debug a failed execution
- docs index