fix(tui): 修复 /help 单页可选执行 + UTF-8 + 覆盖率修复 by creatang · Pull Request #208 · 1024XEngineer/neo-code

creatang · 2026-04-09T14:56:49Z

变更说明

将 /help 改为与 /model 一致的可选择弹层展示
在当前界面直接展示全部 slash 命令，不切换页面
支持在 help 列表中直接选择并执行 slash 命令
/help 列表改为单页展示（关闭分页并自适应高度）

UTF-8 与覆盖率修复

按 UTF-8 更新 PR 标题与描述，修复乱码
补充 update.go、view.go、copy_code.go、app.go 相关测试路径
移除 refreshHelpPicker 无效错误分支，降低不可达覆盖噪音

验证

go test ./internal/tui/...

codecov · 2026-04-09T14:58:18Z

Codecov Report

❌ Patch coverage is 90.73684% with 44 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
internal/tui/core/app/update.go	87.56%	18 Missing and 5 partials ⚠️
internal/provider/errors.go	78.57%	3 Missing and 3 partials ⚠️
internal/provider/openai/events.go	60.00%	3 Missing and 1 partial ⚠️
internal/runtime/compact.go	76.47%	3 Missing and 1 partial ⚠️
internal/provider/openai/response.go	85.71%	1 Missing and 1 partial ⚠️
internal/runtime/permission.go	50.00%	1 Missing and 1 partial ⚠️
internal/tui/core/app/copy_code.go	91.66%	1 Missing and 1 partial ⚠️
internal/runtime/runtime.go	96.15%	1 Missing ⚠️

📢 Thoughts on this report? Let us know!

…-emitter fix(tools): 重构 EmitChunk 错误传播与调用语义

minorcell · 2026-04-10T03:09:50Z

/review

Generated with [codeagent](https://github.com/qbox/codeagent) Co-authored-by: Yumiue <188874804+Yumiue@users.noreply.github.com>

…t, rate-limit misclassification - runtime.go: decrement attempt before reactive-compact continue so the retry does not consume a MaxLoops slot (fixes max_loops=1 case) - compact.go: reset TokenInputTotal/TokenOutputTotal on session before saving after compact, preventing stale high totals from being restored on the next run - errors.go: guard text-based context_too_long override to only apply for generic client errors (ErrorCodeClient); 429 rate-limit errors whose message contains token-count fragments are no longer mis-routed into the reactive-compact path; same guard added in IsContextTooLong - errors_test.go: add regression cases for the 429-with-token-count-message scenarios in NewProviderErrorFromStatus and IsContextTooLong Generated with [codeagent](https://github.com/qbox/codeagent) Co-authored-by: Yumiue <188874804+Yumiue@users.noreply.github.com>

Generated with [codeagent](https://github.com/qbox/codeagent) Co-authored-by: Cai-Tang-www <106404101+Cai-Tang-www@users.noreply.github.com>

Generated with [codeagent](https://github.com/qbox/codeagent) Co-authored-by: Yumiue <188874804+Yumiue@users.noreply.github.com>

…e-146-1775790664 feat(runtime): recover with reactive compact on context-too-long

…-ask-select-closure feat(tui/runtime): 补齐权限 ask 的选择式闭环（上下选择 + Enter确认）

Generated with [codeagent](https://github.com/qbox/codeagent) Co-authored-by: minorcell <120795714+minorcell@users.noreply.github.com>

Generated with [codeagent](https://github.com/qbox/codeagent) Co-authored-by: Cai-Tang-www <106404101+Cai-Tang-www@users.noreply.github.com>

…e-213-1775793526 fix(runtime/tools): 收敛 EmitChunk 遗留 review 问题

Generated with [codeagent](https://github.com/qbox/codeagent) Co-authored-by: minorcell <120795714+minorcell@users.noreply.github.com>

creatang · 2026-04-10T07:25:40Z

/review -codex

fennoai

Found 2 regressions in the current diff: one in OpenAI stream completion/cancellation handling, and one in TUI layout sizing vs. hit-testing.

fennoai · 2026-04-10T07:30:25Z

-		line, err := reader.ReadLine()
+		// 每次读取前优先响应上下文取消，避免取消请求被误判为流中断。
+		select {
+		case <-ctx.Done():


Once done has been set after a [DONE] frame, this unconditional ctx.Done() check can still return context.Canceled before the trailing blank line/EOF is consumed. In that case a successfully completed stream is reported as canceled and message_done is never emitted. A regression test that cancels after [DONE] but before EOF would reproduce this.

fennoai · 2026-04-10T07:30:25Z

-	a.transcript.Height = max(6, lay.rightHeight-activityHeight-menuHeight-promptHeight)
+	availableHeight := lay.rightHeight - activityHeight - menuHeight - promptHeight
+	minTranscriptHeight := max(6, lay.rightHeight/2)
+	a.transcript.Height = max(minTranscriptHeight, availableHeight)


a.transcript.Height now uses max(lay.rightHeight/2, availableHeight), but renderWaterfall() still renders the transcript with max(6, height-activity-menu-prompt). When those diverge, the visual transcript/prompt positions no longer match the coordinates used by inputBounds() / activityBounds(), so mouse hit-testing and scrolling become offset on layouts where the activity pane or command menu is visible.

fennoai

Two issues worth addressing before merge.

fennoai · 2026-04-10T07:31:18Z

-	a.transcript.Height = max(6, lay.rightHeight-activityHeight-menuHeight-promptHeight)
+	availableHeight := lay.rightHeight - activityHeight - menuHeight - promptHeight
+	minTranscriptHeight := max(6, lay.rightHeight/2)
+	a.transcript.Height = max(minTranscriptHeight, availableHeight)


a.transcript.Height is now forced to at least half of the right pane, even when the activity panel, command menu, and prompt leave less space than that. The new renderWaterfall() path recomputes a smaller visible transcript height, so the viewport's internal height diverges from what is actually rendered. On shorter terminals this throws off AtBottom/scroll calculations and can clip or jump the transcript. The viewport height should stay in sync with the rendered container instead of enforcing a larger minimum here.

fennoai · 2026-04-10T07:31:18Z

 			autoCompacted = true
 			var compactResult contextcompact.Result
-			session, compactResult, _ = s.runCompactForSession(ctx, input.RunID, session, cfg, false)
+			session, compactResult, _ = s.runCompactForSession(ctx, input.RunID, session, cfg, contextcompact.ModeManual, false)


This is the new token-threshold auto-compact branch, but it still calls runCompactForSession(..., contextcompact.ModeManual, ...). As a result every compact_* event from an automatic compact is reported as manual, so the UI/docs added in this PR cannot distinguish an auto-compact from an explicit /compact. If auto-compaction is now a first-class path, it needs its own trigger mode instead of reusing manual.

fennoai

Two non-blocking issues remain in the changed code paths after verification.

fennoai · 2026-04-10T07:31:58Z

+	"maximum context length",
+	"maximum prompt length",
+	"prompt is too long",
+	"requested too many tokens",


"requested too many tokens" / "too many tokens" are too broad for context_too_long detection. Providers also use those phrases for bad-request cases like oversized max_tokens / requested output, which compaction cannot fix. In those cases this PR will route the error into the new reactive-compact retry path instead of surfacing the real configuration problem.

fennoai · 2026-04-10T07:31:59Z

+	activityHeight := a.activityPreviewHeight()
+	menuHeight := a.commandMenuHeight(width)
+	promptHeight := lipgloss.Height(a.renderPrompt(width))
+	transcriptHeight := max(6, height-activityHeight-menuHeight-promptHeight)


renderWaterfall now recomputes its own transcript height instead of using a.transcript.Height, but applyComponentLayout, inputBounds, and transcriptBounds still use a.transcript.Height. On short panes those values diverge (applyComponentLayout enforces a half-pane minimum, this code does not), which will misplace the rendered prompt/activity area relative to the viewport and break mouse hit-testing/scroll positioning.

fennoai

Found 2 regressions worth fixing before merge.

fennoai · 2026-04-10T07:32:09Z

-		line, err := reader.ReadLine()
+		// 每次读取前优先响应上下文取消，避免取消请求被误判为流中断。
+		select {
+		case <-ctx.Done():


Once [DONE] has been seen, this pre-read ctx.Done() check can still turn a successfully completed stream into context.Canceled if the request is canceled before the trailing blank line/EOF is consumed. In that case message_done never fires even though the model already finished. The cancel fast-path should be skipped after done == true, or handled after the pending completion is flushed.

fennoai · 2026-04-10T07:32:09Z

@@ -251,12 +252,12 @@ func (s *Service) Run(ctx context.Context, input UserInput) error {
 		if builtContext.ShouldAutoCompact && !autoCompacted {
 			autoCompacted = true


autoCompacted is flipped before we know whether compacting actually succeeded. Because runCompactForSession(..., failOnError=false) swallows runner/save errors and can also return Applied=false, a failed or no-op auto-compact permanently disables further threshold-triggered compaction for the rest of the run. That leaves later oversized turns with no second chance to compact.

fennoai

A few regressions still need attention before this lands. The main issues are stream completion being converted into cancellation, overly broad context-too-long detection, and the new transcript sizing getting out of sync with the viewport model.

fennoai · 2026-04-10T07:33:41Z

-		line, err := reader.ReadLine()
+		// 每次读取前优先响应上下文取消，避免取消请求被误判为流中断。
+		select {
+		case <-ctx.Done():


If the provider has already sent [DONE], canceling the request before the trailing blank line / EOF now returns context.Canceled here instead of finishing the stream. That skips message_done, so the last assistant turn and token usage can be dropped during normal shutdown races. Once done is true, this pre-read cancel check should stop treating the stream as incomplete.

fennoai · 2026-04-10T07:33:41Z

+	"maximum context length",
+	"maximum prompt length",
+	"prompt is too long",
+	"requested too many tokens",


The new "requested too many tokens" / "too many tokens" fragments are broad enough to catch ordinary 400 validation errors that are not caused by prompt/context size (for example, asking for too many completion tokens). Those requests will now be upgraded to context_too_long and sent through the reactive-compact retry path even though compact cannot fix them.

fennoai · 2026-04-10T07:33:41Z

-	a.transcript.Height = max(6, lay.rightHeight-activityHeight-menuHeight-promptHeight)
+	availableHeight := lay.rightHeight - activityHeight - menuHeight - promptHeight
+	minTranscriptHeight := max(6, lay.rightHeight/2)
+	a.transcript.Height = max(minTranscriptHeight, availableHeight)


a.transcript.Height now uses max(minTranscriptHeight, availableHeight), but renderWaterfall still renders the transcript with just availableHeight. On short panes the viewport model thinks it is taller than the box actually drawn on screen, which throws off bottom detection/scroll math and causes clipping or jumpy scrolling.

creatang · 2026-04-10T07:41:49Z

@xgopilot 解决review提到的问题

fennoai · 2026-04-10T07:42:00Z

已完成

已修复 review 提到的几处回归，并已推送到当前 PR 分支，提交为 3fe66f0。

本次处理

修复 OpenAI SSE 在收到 [DONE] 后遇到取消竞争时误返回 context canceled，保证 message_done 仍能收尾发出。
为阈值触发的 compact 增加独立 auto 模式，并修正只有在 auto compact 真正生效后才标记已触发，避免 no-op/失败后屏蔽后续 reactive compact。
收紧 context_too_long 文本匹配，避免把 max_tokens 之类的普通 token 校验错误误归类到 compact 重试路径。
统一 TUI waterfall/transcript 高度计算，消除渲染高度与 viewport / 鼠标命中区域不一致的问题。
补充对应回归测试，覆盖 DONE 后取消、auto compact 模式与 no-op 退化路径、provider 错误分类、TUI 布局同步。

验证

go test ./internal/context/compact ./internal/provider ./internal/provider/openai ./internal/runtime ./internal/tui/core/app
go test ./...

View job run

Generated with [codeagent](https://github.com/qbox/codeagent) Co-authored-by: creatang <165447160+creatang@users.noreply.github.com>

Cai-Tang-www and others added 6 commits April 9, 2026 17:10

fix(tools): 重构EmitChunk错误传播与并发语义

560a716

fix: 修复流取消误判并补齐覆盖用例

05ad184

fix(test):补充测试覆盖率

b574036

fix：代码块渲染

c31d7e9

fix（tui）：固定输入区

6f9dd4c

refactor:美化/help界面

0e4af34

fix(tui): 修复 /help 编码与覆盖率问题（UTF-8）

55c08a4

creatang changed the title ~~refactor: ?? /help ???????????~~ fix(tui): /help ?????? + UTF-8 + ????? Apr 9, 2026

creatang changed the title ~~fix(tui): /help ?????? + UTF-8 + ?????~~ fix(tui): 修复 /help 单页可选执行 + UTF-8 + 覆盖率修复 Apr 9, 2026

creatang and others added 4 commits April 9, 2026 23:27

test(tui):补充 update 分支覆盖，修复 codecov patch 覆盖率

8ca48f0

Merge pull request 1024XEngineer#205 from Cai-Tang-www/fix/emit-chunk…

64fb3b1

…-emitter fix(tools): 重构 EmitChunk 错误传播与调用语义

feat(tui/runtime): 打通权限审批选择菜单闭环并增强ask提示词

6b5a253

test(tui): 补齐权限审批闭环分支覆盖并修复布局边界

29b9702

This comment was marked as outdated.

Sign in to view

xgopilot and others added 10 commits April 10, 2026 03:17

feat(runtime): recover with reactive compact on context-too-long

14ecb63

Generated with [codeagent](https://github.com/qbox/codeagent) Co-authored-by: Yumiue <188874804+Yumiue@users.noreply.github.com>

fix(tui): close permission ask review gaps

c7dc253

Generated with [codeagent](https://github.com/qbox/codeagent) Co-authored-by: Cai-Tang-www <106404101+Cai-Tang-www@users.noreply.github.com>

fix(runtime): rebuild context after auto compact

825f225

Generated with [codeagent](https://github.com/qbox/codeagent) Co-authored-by: Yumiue <188874804+Yumiue@users.noreply.github.com>

Merge pull request 1024XEngineer#211 from 1024XEngineer/xgopilot/issu…

654adbe

…e-146-1775790664 feat(runtime): recover with reactive compact on context-too-long

Merge pull request 1024XEngineer#210 from Cai-Tang-www/fix/permission…

8c1f663

…-ask-select-closure feat(tui/runtime): 补齐权限 ask 的选择式闭环（上下选择 + Enter确认）

fix(tui): address help modal and review regressions

b381864

Generated with [codeagent](https://github.com/qbox/codeagent) Co-authored-by: minorcell <120795714+minorcell@users.noreply.github.com>

fix(runtime/tools): 收敛 EmitChunk 遗留 review 问题

45294f5

Generated with [codeagent](https://github.com/qbox/codeagent) Co-authored-by: Cai-Tang-www <106404101+Cai-Tang-www@users.noreply.github.com>

fix(runtime): resolve unresolved EmitChunk reviews

a4a2749

Generated with [codeagent](https://github.com/qbox/codeagent) Co-authored-by: Cai-Tang-www <106404101+Cai-Tang-www@users.noreply.github.com>

Merge pull request 1024XEngineer#215 from 1024XEngineer/xgopilot/issu…

1f7eea9

…e-213-1775793526 fix(runtime/tools): 收敛 EmitChunk 遗留 review 问题

fix(tui): resolve pr-208 conflicts with main

8b44672

Generated with [codeagent](https://github.com/qbox/codeagent) Co-authored-by: minorcell <120795714+minorcell@users.noreply.github.com>

1024XEngineer deleted a comment from fennoai Bot Apr 10, 2026

creatang added 8 commits April 10, 2026 15:07

fix(test):补充测试覆盖率

5736244

fix：代码块渲染

f947e2f

fix（tui）：固定输入区

6c92d04

refactor:美化/help界面

e9b490d

fix(tui): 修复 /help 编码与覆盖率问题（UTF-8）

760fa46

test(tui):补充 update 分支覆盖，修复 codecov patch 覆盖率

90391e1

fix(test): rebase后对齐权限提示测试

aa1e6d1

merge: resolve conflicts with remote branch

144e499

fennoai Bot reviewed Apr 10, 2026

View reviewed changes

fix: address PR review regressions

3fe66f0

Generated with [codeagent](https://github.com/qbox/codeagent) Co-authored-by: creatang <165447160+creatang@users.noreply.github.com>

minorcell merged commit 0e60d38 into 1024XEngineer:main Apr 10, 2026
2 checks passed

		@@ -251,12 +252,12 @@ func (s *Service) Run(ctx context.Context, input UserInput) error {
		if builtContext.ShouldAutoCompact && !autoCompacted {
		autoCompacted = true

Uh oh!

Conversation

creatang commented Apr 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

变更说明

UTF-8 与覆盖率修复

验证

Uh oh!

codecov Bot commented Apr 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

minorcell commented Apr 10, 2026

Uh oh!

This comment was marked as outdated.

Uh oh!

This comment was marked as outdated.

Uh oh!

This comment was marked as outdated.

Uh oh!

This comment was marked as outdated.

Uh oh!

This comment was marked as outdated.

Uh oh!

creatang commented Apr 10, 2026

Uh oh!

fennoai Bot left a comment

Choose a reason for hiding this comment

Uh oh!

fennoai Bot Apr 10, 2026

Choose a reason for hiding this comment

Uh oh!

fennoai Bot Apr 10, 2026

Choose a reason for hiding this comment

Uh oh!

fennoai Bot left a comment

Choose a reason for hiding this comment

Uh oh!

fennoai Bot Apr 10, 2026

Choose a reason for hiding this comment

Uh oh!

fennoai Bot Apr 10, 2026

Choose a reason for hiding this comment

Uh oh!

fennoai Bot left a comment

Choose a reason for hiding this comment

Uh oh!

fennoai Bot Apr 10, 2026

Choose a reason for hiding this comment

Uh oh!

fennoai Bot Apr 10, 2026

Choose a reason for hiding this comment

Uh oh!

fennoai Bot left a comment

Choose a reason for hiding this comment

Uh oh!

fennoai Bot Apr 10, 2026

Choose a reason for hiding this comment

Uh oh!

fennoai Bot Apr 10, 2026

Choose a reason for hiding this comment

Uh oh!

fennoai Bot left a comment

Choose a reason for hiding this comment

Uh oh!

fennoai Bot Apr 10, 2026

Choose a reason for hiding this comment

Uh oh!

fennoai Bot Apr 10, 2026

Choose a reason for hiding this comment

Uh oh!

fennoai Bot Apr 10, 2026

Choose a reason for hiding this comment

Uh oh!

creatang commented Apr 10, 2026

Uh oh!

fennoai Bot commented Apr 10, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

已完成

本次处理

creatang commented Apr 9, 2026 •

edited

Loading

codecov Bot commented Apr 9, 2026 •

edited

Loading

fennoai Bot commented Apr 10, 2026 •

edited

Loading