Repository navigation
Preserved branch: queue/3201-2167-shebang-dispatch-tmp — decision needed (merge / cherry-pick / drop) #3596
Description
Preserved branch from the laptop worktree cleanup of 2026-11-01.
- Branch:
queue/3201-2167-shebang-dispatch-tmp - Verified HEAD on origin:
52dbf17d6516eaf99bc8d7d2fe667a274edd6521 - Provenance: this branch holds work that existed only in a local worktree on the laptop (unique commits and/or a rescue commit of uncommitted work: "rescue: preserve uncommitted work from worktree cleanup 2026-11-01"). It was pushed so the worktree could be removed without losing the work; the local tree and local branch are gone.
- Decision needed: merge it, cherry-pick from it, or drop the branch. Nothing else references it; this issue is the record of what it holds.
Activity
developing-today commentedon Nov 1, 2026
CODE-1423
dezren39 commentedon Nov 3, 2026
agent-note:v1 — lane
335a5d1f-bfc1-4831-99ea-850c8939a727(repo-worker-code), task 7216ingested → verified → closing (already landed)
Verification against current
main(fetched 2026-11-03):queue/3201-2167-shebang-dispatch-tmp(HEAD52dbf17d6) holds 9 commits of the mcpx: shebang-first interpreter dispatch + script.interpreter + script.cwd (split of #1919) #2167 shebang-dispatch work, all dated 2026-10-30, and its last commit's own comments cite "mcpx: shebang-first interpreter dispatch + script.interpreter + script.cwd (#2167) #2611 review".8e7b50492.dispatch_test.go: every one of the 28 test functions on the branch exists onmain;mainhas 36 (8 later additions).dispatch.go:maincontains the branch's later-iteration features (env terminator assignments, raw-Spayload,-u/-i/-C) plus refinements the branch never got —SplitEnvStringwith env(1) escape semantics andoptErrrefusal of unsupported env options — and the branch still calls the olderSplitCommandLineinresplit.main.main'spkgs/mcpx/internal/runner/dispatch.gohas had no functional change since that merge (only a later merge commit in its history).Decision: dropped as obsolete — the work landed via PR #2611. Nothing to cherry-pick; reopening this as a PR would regress
dispatch.goto the earlier parser. Per the brief, the branch itself is not deleted (left as the historical record; it can be removed whenever the fleet retires preserved refs).Closing this issue with this evidence.