Commit Graph
8 Commits
Author SHA1 Message Date
liukx0205 b263581e92 fix(client): answer the review bot's findings on the keyboard hand-offs
- The model seat decides before consuming: a focused control that is not a row
  (a retry button) keeps the browser's traversal, and Tab enters the list only
  from the trigger.
- A launcher-opened menu is a fresh intent, so the dismissal that closed the
  same token earlier no longer silences it.
- Comments and JSDoc now match the implementation: the near-end entry resumes
  from the walk's last index, Shift+Tab leaves whenever the menu is open, and
  the anchor hand-back names the control that opened the menu.
- The e2e focus assertions all go through `expect.poll`.
- Regression coverage for a rowless pane (the trigger keeps the keyboard) and
  for a focused retry keeping its Tab, in the seat and the popup shell.
- The PR records the two public-surface changes (`SessionInput.focus()` in,
  `bindComposerFocus` out) and the row-fill focus decision.
2026-09-15 14:21:00 +08:00
liukx0205 c47a469aec feat(client): accept and leave with Enter, Tab, and Escape in the composer
The popup select panel and the model seat hold focus while open, and their Tab
fell through to the browser's traversal, leaving an open surface that only a
pointer press clears. Both now settle on Enter or Tab, leave on Escape or
Shift+Tab, walk with the arrows, and park on the value in use: the panel's
highlight on the row its options mark current, the seat's drilled pane on its
checked row, with the keyboard handed back to the cell that drilled in when the
pane is left.

Dismissal rides the conversation's own focus(), so the caret returns where the
draft left it instead of at the start.
2026-09-15 13:35:04 +08:00
imccyu 9135a13a8b refactor(consumers): remove cross-package runtime relays 2026-08-30 02:29:52 +08:00
Tianyi Cui ce031ddd16 fix(web): cover authenticated host runtimes 2026-08-25 14:23:46 +08:00
Tianyi Cui caf386f59c fix(test): use expected-output naming 2026-08-24 21:37:14 +08:00
Tianyi Cui 6189e4a374 test(web): separate session snapshots from goldens 2026-08-24 21:36:31 +08:00
Tianyi Cui 9d5eb37638 fix(headless): dsh run is a direct core front door 2026-08-09 13:44:44 +08:00
Yichen Jiang 0490f8bb06 feat(llm-pi-ai): per-model reasoningEfforts and reasoning-dispatch compat switches
A model entry's reasoningEfforts dict declares its selectable thinking
levels — key = offered level, value = the wire spelling dispatch sends;
only off may leave the value empty (supported, send nothing). false
strips reasoning from a catalog model; every level is materialized
explicitly into pi-ai's thinkingLevelMap so nobody has to know pi-ai's
asymmetric absent-key defaulting. compat.thinkingFormat and
compat.supportsReasoningEffort become configurable on the route and per
model (model > route > catalog entry > pi-ai's URL-derived guess),
openai-completions only, so a private gateway speaking the DeepSeek
reasoning dialect no longer depends on its URL being recognizable.
Record-typed drift gates pin both enums to pi-ai's, and an unserviceable
declaration is refused at the write that produced it, naming route,
model, and level.
2026-08-08 00:51:54 +08:00