ci(release): add a manual dry run that builds everything and publishes nothing - #14

…s and the macOS binary (#15)

## Summary

The release page of a tag shows only the `CHANGELOG.md` section that has the version of the tag (`release.yml` extracts it with `awk`). Today that section, `[0.5.0]`, has nothing about the work in `tui/`, and `[Unreleased]` has most of it. This PR makes `[0.5.0]` the text of the release. It is exit condition 5 of the V0.5 milestone ([v0.5.0](https://github.com/zy1233/pi-python/milestone/1)).

Three files change: `CHANGELOG.md`, and, for the macOS binary, `README.md` and `docs/TUI-AND-CODE-AGENT.md`. The `CHANGELOG.md` diff is larger than the new text, because the moved bullets show as removed and added.

- **Fold.** The 18 bullets of `[Unreleased]` (9 Added, 5 Changed, 4 Fixed) move into the matching groups of `[0.5.0]`, after the bullets that are already there. `[Unreleased]` is empty. The date of `[0.5.0]` is 2026-10-12.
- **A summary paragraph** at the top of `[0.5.0]`: what the release is.
- **20 new bullets** for things that were in neither section:

| Group | New entries |
|---|---|
| Added | `/model` through ACP session config options; **a prebuilt `zypi` for macOS on Apple silicon** |
| Changed | `zypi -p` refuses a sandbox and names the flags it ignores; `zypi` is the product name where a name is shown |
| Removed (new group) | the Rust agent runtime (with the numbers); the screens, commands and flags that went with it |
| Fixed | a tool left running after quit; `session/close` and `session/resume`; the title and time in `session/list`; `AGENTS.md` read twice on macOS; the `ssh-wrap` alias; messages that name things that do not exist |
| Known issues (new group) | 8 items, from the list in the plan |

- **The macOS binary** (the second commit; you chose this after the dry run in #14 passed). The Added entry, the Known issue "tested on Linux only", the install steps in the TUI guide, the platform table, the download paragraph of the README and its roadmap line say that `zypi-macos-arm64` exists, that CI builds it, and that nobody has started it.

## Where the facts come from

- **2,500 to 1,755 `.rs` files, 1,620,612 to 1,021,067 lines** under `tui/crates`: `git grep -c -e "^"` at `v0.4.0` and at `main` (`4fc028a`).
- **1,311 to 1,257 packages**: the `[[package]]` entries of `tui/Cargo.lock` at the same two revisions.
- **470,461,776 to 420,767,096 bytes** (−10.6 %): the `release_baseline` run [38221952524](https://github.com/zy1233/pi-python/actions/runs/38221952524) on `4fc028a`. The same run builds in 1,195 s (v0.4.0: 1,647 s) and `--version` takes 23.4 ms.
- **Removed flags** `--fork-session`, `--leader`, `--no-leader` and `--leader-socket`: I compared the long flags of `PagerArgs` (`pi-pager/src/app/cli.rs`) at `v0.4.0` and `main`. 7 flags were removed and none was added. The other 3 are the `--local-workspace*` flags, which a release build never had, so they are not in the entry.
- **The macOS binary**: the dry run [38223353755](https://github.com/zy1233/pi-python/actions/runs/38223353755) (from the branch of #14) built `zypi-macos-arm64` on `macos-14` in 25 min. `release.yml` has no signing or notarization step.
- **The other entries** come from the code and from sections 10.2, 10.7, 10.8 and 10.10 of `docs/PLAN/archive/PLAN-RUST-AGENT-RUNTIME-REMOVAL-EXECUTION-LOG.md`. I read the code again for `-p` (`print_mode.rs`), the `session/list` title (`session_list.py`), the `AGENTS.md` check (`context_files.py`) and the `session/close` flag (`__main__.py`).

## Decisions for you

1. **The date.** It is 2026-10-12 (today). If you tag on another day, change it before the tag.
2. **Known issues in the release text.** The release page will show the 8 items. They are true as far as I know, but they are your public statement. Cut any you do not want to publish. One of them (`Version numbers`) goes away with #17 (plan decision 16). #17 is stacked on this PR and merges right after it.

## Not verified

- How GitHub shows the text. I ran the `awk` program of `release.yml` (Git for Windows `awk`) on this file: the body is 44 KB, has the 5 groups, and has no `## [` line.
- The version claim in Known issues (`zypi --version` shows `1.0.8`) comes from the code (`build.rs`, `Cargo.toml`). I did not run a release binary when I wrote this.
- Nobody has run the macOS binary. The advice about `xattr -d com.apple.quarantine` is general macOS behaviour. I did not test it.
- No person other than me has read the new English text.

## Verification

- Bullet lines before: 75. After: 95. The difference is the 20 new bullets. No bullet was lost or changed except the Known issue "tested on Linux only", which has one more sentence. The text from `## [0.4.0]` on is the same.
- Every backtick and `**` pair in `CHANGELOG.md` is balanced.
- `ruff check .` and `ruff format --check .`: clean (nothing in Python changes).

…(plan r12)

Define what V0.5 is from the state of the Rust agent runtime removal on
main (4fc028a; CI and TUI CI are green on it): the untagged 0.5.0 plus
the TUI side of r3-r11. The scope follows the status table. The plan now
holds the six exit conditions, the known issues that ship with the
release and the items that do not ship. The evidence is in section 10.15
of the archived execution log.

Found while checking the release path:
- release.yml had no manual trigger, and its macOS arm64 job had never
  run in a release (PR #14 adds the trigger);
- the CHANGELOG had no entry for the removal itself (PR #15 adds it);
- zypi --version and "is this a release build" share GROK_VERSION, so the
  obvious fix for the version string would turn folder trust on (new plan
  decision 16).

The follow-ups that were chosen are recorded here, not done here:
- release size (condition 2): 420,767,096 B, 10.6 % below v0.4.0, from the
  release_baseline run 38221952524; docs/baselines/tui.md has the numbers;
- dry run (condition 3): run 38223353755, started from the PR branch,
  built Linux x86_64 and macOS arm64 and published nothing;
- release notes (condition 5): PR #15;
- decision 16: option (a) chosen, not implemented.
The comparison with v0.4.0 (condition 4) is not done. WSL does not start on
this machine, so no release binary was run.

A GitHub milestone v0.5.0 was created next to this change. Nothing is
attached to it. Nothing under tui/, packages/ or .github/ changes.

…L2 (plan r12, part 3)

#15, #17 and #14 are squash-merged (59545c1, d57fd9b, 0efd400). The plan and the
execution log say so, and main is now 0efd400 in the current-state text.

WSL2 works again. On 0efd400: Linux pytest 2,539 passed / 16 skipped / 31
deselected; `zypi --version` prints `zypi 0.5.0 (0efd400)` (a debug build
on ext4, not the CI release-dist artifact); the PTY scripts pass, except one
check of smoke.py, which is out of date since #12 (22/23). Section 10.15 gets a
part 3 with these results and the leftovers; the "not verified" lines that said
"WSL does not start" are brought up to date.

Only documents change.

…(plan r12) (#16)

## Summary

This PR records the V0.5 release in the plan: what the release is, what must be true before the tag, and what is known to be wrong. It is revision r12 of the plan for the removal of the Rust agent runtime. Only documents change.

The milestone is on GitHub: [v0.5.0](https://github.com/zy1233/pi-python/milestone/1). Nothing is attached to it, because the conditions are written in one place (the plan).

The three PRs that do the work of the conditions are merged before this one: #15 (`59545c1`), #17 (`d57fd9b`) and #14 (`0efd400`), in that order, each as one squash commit. This branch is rebased onto `0efd400`.

- `docs/PLAN/PLAN-RUST-AGENT-RUNTIME-REMOVAL.md`: a new section "发布里程碑 V0.5" (what the release is, six exit conditions, known issues, what is not in it). The status, the next steps and the "not verified" list start from `main` at `0efd400`. Decision 16 is new, and #17 did it.
- `docs/PLAN/archive/PLAN-RUST-AGENT-RUNTIME-REMOVAL-EXECUTION-LOG.md`: section 10.15, with the reasons, the figures, the follow-up work, your three answers (part "续 2"), and the merges and the WSL2 checks (part "续 3").
- `docs/baselines/tui.md`: the numbers of the new release measurement.

## The six exit conditions

| # | Condition | State |
|---|---|---|
| 1 | CI is green on the commit that gets the tag | Green: `CI` on `0efd400`, `TUI CI` on `d57fd9b` (`0efd400` differs from it only in `release.yml`). Look again on the commit that gets the tag. |
| 2 | The Linux release binary is not larger than v0.4.0 | **Done.** 420,767,096 B (v0.4.0: 470,461,776 B, 10.6 % smaller). Run [38221952524](https://github.com/zy1233/pi-python/actions/runs/38221952524). |
| 3 | A dry run of the release build on Linux and macOS arm64 | **Done.** Run [38223353755](https://github.com/zy1233/pi-python/actions/runs/38223353755) built both and published nothing (27 min). The manual trigger is in #14, merged. |
| 4 | The PTY scripts on the v0.4.0 binary and on the new binary | **Skipped by you.** V0.5 does not wait for it. Which known issues come from v0.4.0 stays a guess. |
| 5 | The release notes | **Done** in #15, merged. They list the macOS binary and say that CI built it and nobody has started it. #17 added one entry and removed one known issue. |
| 6 | The tag `v0.5.0` | Yours. |

## After the merges, with WSL2 working again

You fixed WSL. These runs are on `0efd400` (details: section 10.15, part "续 3").

- Linux `pytest`: 2,539 passed / 16 skipped / 31 deselected.
- `zypi --version` prints `zypi 0.5.0 (0efd400)`, exit 0. That is a debug build on ext4. It is not the `release-dist` file that CI builds.
- `agentInfo.version` is `0.5.0` in a venv installed from the merged tree. An older editable install keeps the version it had when it was installed (the old `/tmp/wsl-venv`: 0.2.0).
- PTY scripts: `exit_matrix.py` 12/12, `print_exit.py` 4/4, `resume_matrix.py` 10 PASS / 3 GAP (the known gaps), `mode_probe.py` as in #12, `status_line_probe.py` 4/4, `net_audit.py --mock` 13/13 and with `--no-network` 13/13.
- **`smoke.py` is 22/23.** The failing check, "status line follows the switch", waits for the text `Second model` followed by a `·`. The screen shows the right label, `Second model`, the session file has the `model_change`, and the next turn answers. The check is out of date since #12. My inference, not tested on an old build: the `·` came from the `auto` flag, which #12 no longer shows in the normal mode. It is a script problem, not a TUI problem. It is not fixed here. The fix is one regular expression.

## Decisions for you

- The plan decisions 6, 8, 14 and 15 should be made before the tag. If they are not, they go in the release as known issues.
- The date of `[0.5.0]` in the changelog is 2026-10-12. Change it on the day of the tag. The known issues become public release text.

## Not verified

- Nobody ran the `release-dist` binary that CI builds; only a debug build of `main` ran. Nobody ran the macOS binary. The changelog, the README and the TUI guide say so.
- The two changed `if` lines of `release.yml` have run only on their false side. A tag push runs the true side.
- This round did not run the Landlock layer (kernel 5.10 has none), `smoke.py --sandbox`, or the whole Rust test manifest on WSL2. CI ran the manifest on `d57fd9b`: 8,665 passed / 0 failed / 22 ignored.
- The plan says that the version overrides of `pi-config` could stop working for a user config that needs a version above `0.5.0`. I found no default config that does, and I cannot see user configs.

## Verification

- Link check of the plan and of the execution log: plan 149 local and 81 external links, 47 reference definitions; log 43 local and 58 external links; 0 unused, 0 problems.
- `ruff check .` and `ruff format --check .`: clean (nothing in Python changes).
- The test counts, run ids and times come from the job logs (`gh run view`) and from the output of the runs on WSL2.
- The three merges used `--match-head-commit`. #15 and #17 merged with the tree that CI had tested (#17 was rebased onto `main` after #15, and `git diff` against its tested tip was empty). #14 merged onto a newer `main`, so `CI` ran again on the result, `0efd400`, and passed.