| 1 | # Current lane state |
| 2 | |
| 3 | **This file is expected to go stale. Verify before you act on it.** |
| 4 | |
| 5 | Everything here is a snapshot of a moving release lane. `AGENTS.md` holds the |
| 6 | durable rules; this holds the perishable state, so that a stale milestone number |
| 7 | never sits next to a rule and borrows its authority. |
| 8 | |
| 9 | Last updated: 2026-08-02. |
| 10 | |
| 11 | ## Lane |
| 12 | |
| 13 | - **Repo:** `Hmbown/CodeWhale`. It lives on multiple devices — work in whichever |
| 14 | local checkout you have and confirm with `git branch --show-current` before |
| 15 | editing. |
| 16 | - **Branch:** recent work lands on `main` through small PRs rather than a |
| 17 | long-lived `codex/...` integration branch. Verify any named integration branch |
| 18 | still exists before relying on it. |
| 19 | - **Workspace version:** `0.9.4` — but read it from `Cargo.toml` |
| 20 | (`[workspace.package] version`), which is the source of truth over any number |
| 21 | written here. |
| 22 | - **Milestone:** `v0.9.4`. List it live rather than trusting this line: |
| 23 | |
| 24 | ```sh |
| 25 | gh issue list --repo Hmbown/CodeWhale --milestone "v0.9.4" --state open |
| 26 | ``` |
| 27 | |
| 28 | ## Known test flakes (pre-existing, not regressions) |
| 29 | |
| 30 | - `run_verifiers_background_*` flakes under full-suite parallelism but passes in |
| 31 | isolation. Rerun in isolation before blaming your change. |
| 32 | |
| 33 | ## Closed investigations — do not reopen without new evidence |
| 34 | |
| 35 | - **Sub-agent TUI freeze** (reported in older handoffs) is resolved by the |
| 36 | v0.8.61 cutover: cap-20, persist-debounce, AgentProgress redraw throttle, |
| 37 | ListSubAgents coalescing, input-pump-off-render-thread. The leading "blocking |
| 38 | I/O starves the worker pool" theory was measured and **disproven** |
| 39 | (`git rev-parse` ~10ms, 18-core machine). Do not spend effort on a speculative |
| 40 | `spawn_blocking` fix. |
| 41 | - The old `config_command_allow_shell_*` failures on machines with |
| 42 | `default_mode = "yolo"` were fixed by pinning the command-test app to Agent |
| 43 | mode. |
| 44 | |
| 45 | ## Known debt (deliberate, not a bug to fix casually) |
| 46 | |
| 47 | - The workflow *history* card renders with `Locale::En` until locale is threaded |
| 48 | through `ToolCell::lines_with_mode` (~30 call sites). |
| 49 | - The `classic` ocean treatment exists in code but persisted settings normalize |
| 50 | it away; do not expand it without a product decision. |
| 51 |