tick-history: 2026-04-29T05:11Z — fourteenth tick narrow drain

docs/hygiene-history/ticks/2026/04/29/0511Z.md

@@ -0,0 +1 @@
+| 2026-04-29T05:11:35Z | claude-opus-4-7 | 26f978a2 | fourteenth tick — narrow drain. Two thread fixes: (a) PR #775 Codex P2 — the "with a local main tracking origin" example used `git switch main` (bare), which only works AFTER `refs/heads/main` already exists locally; on first-run scripts where the local branch hasn't been created yet, bare `main` would still hit the multi-remote ambiguity fatal. Fixed by replacing bare `git switch main` with explicit `git switch refs/heads/main` (unambiguous regardless of local state — asks Git for the local-branch ref directly, never the remote-tracking namespace). Also replaced `git branch --track ... 2>/dev/null \|\| true` with `git branch -f main refs/remotes/origin/main` for unconditional idempotent local ref creation at the explicit start-point. (b) PR #781 same display-artifact-thread class as PR #760 + #777 (Copilot reads "leading space and `1 \|\|` prefix" from a renderer view-mode confusion, not file content); resolved as not-actionable with explanatory comment. | [#775](https://github.com/Lucent-Financial-Group/Zeta/pull/775) [#781](https://github.com/Lucent-Financial-Group/Zeta/pull/781) | Trajectory observation: the bare-main rule has now had THREE rounds of refinement (Amara round-1 + Amara round-2 + Codex round-3 catching the bare-`switch main` slip). Each round caught a real edge case: round-1 = ambiguity + hard-stop; round-2 = orchestrator-vs-shell semantics + concurrency + future-consolidation; round-3 = first-run-no-local-branch edge case. The rule now covers a much larger set of scenarios than the original draft. The seven-step meta-loop fired three times on this single rule, and each iteration tightened the discipline. This is the convergence pattern Amara predicted — multi-AI review catching the things one author misses, even when the author is being careful. |