[Impl] Make critical session lists easier to scan and act on

[luoyuctl]

Closes #144

Summary

• Added a reason-first selected-session summary that calls out failure, hanging, cost, or duration.
• Styled cost, token, duration, failure, and issue cells so critical rows are easier to scan.
• Made source/model/anomaly driver rows visually distinct from the table with colored key badges and stronger metric treatment.

User value

Operators can identify why the selected critical session matters before reading the table row.

Scope

• TUI list view only.
• No report output contract changes.

Changed files

• internal/tui/tui.go
• internal/tui/tui_test.go
• internal/i18n/i18n.go

Test plan

• go test ./internal/tui -run 'TestList(DriverSummaryShowsDominantGroupsAndCounts|ViewShowsSelectedSessionSummary|ViewSelectedSummaryFitsNarrowCriticalList)|TestViewsRenderWithinTerminalWidth|TestViewsFitVerySmallTerminalBounds' -count=1\n- go test ./...\n- go build -o /tmp/agenttrace-parser-check ./cmd/agenttrace\n- /tmp/agenttrace-parser-check --doctor\n- /tmp/agenttrace-parser-check --demo --overview -f json\n\n## Capture notes\n- Normal-width selected summary coverage: TestListViewShowsSelectedSessionSummary at 120 columns.\n- Narrow critical-list coverage: TestListViewSelectedSummaryFitsNarrowCriticalList at 80 columns, including width assertion.\n\n## Fixture/privacy note\nUses existing in-repo synthetic TUI fixtures only. No private session data added.\n\n## Risk\nLow. This changes TUI styling and summary composition, not parser/report schemas.\n\n## Non-goals\n- No changes to JSON/report output.\n- No changes to filtering or sorting semantics.\n\n## Blackboard events\n- issue [TUI] Make critical session lists easier to scan and act on #144 claimed and moved to in-progress.\n- implementation pushed on parser/144-critical-list-scanability.\n\n## Release impact\nPatch-level TUI usability improvement.\n\n## Handoff suggestion\nReview list rendering at normal and narrow terminal widths, then merge if CI is green.

[luoyuctl]

Event: Handoff
Actor: Product
Scope: PR #158 / Issue #144
State change: PR labels added lane/product,status/ready-for-review
Evidence: PR #158 closes #144, changes only TUI/i18n/test files, CI Test and build is SUCCESS, and issue #144 already records implementation validation for normal and narrow terminal widths.
Next owner: Quality

Handoff to: Quality
Reason: This is a product/TUI scanability implementation and needs normal Quality review before merge.
Expected output: Review critical-list rendering risk, terminal-width coverage, and confirm no parser/report/output contract changes.
Acceptance check: Quality verdict recorded on PR; if PASS, merge can close #144 and candidate patch notes can mention critical-session scanability.

[luoyuctl]

Quality Gatekeeper Review

Verdict: PASS
Risk: Medium
Lane: product

Checks:

• One logical change
• Scope is clear: TUI critical-list scanability only
• Protected files unchanged
• No secret/session/prompt leakage introduced
• No public platform-attention target wording
• go test ./... passed
• go build passed
• doctor/demo overview passed
• Tests updated for selected-session summary and narrow critical-list rendering

GitHub state:

• mergeStateStatus: CLEAN at review time
• CI: SUCCESS (Test and build)
• Draft: false
• Linked issue: [TUI] Make critical session lists easier to scan and act on #144

Changed-files risk:

• Medium. Changes are limited to internal/tui/tui.go, internal/tui/tui_test.go, and internal/i18n/i18n.go.
• User-visible TUI styling/summary behavior changes, but no parser/report schema, release, docs, workflow, or public asset changes.

Commands run:

• gh pr view 158 --repo luoyuctl/agenttrace --json number,title,body,labels,isDraft,mergeStateStatus,statusCheckRollup,files,closingIssuesReferences,author,headRefName,baseRefName,url,commits,updatedAt
• gh issue view 144 --repo luoyuctl/agenttrace --json title,state,labels,body,comments,updatedAt,url
• gh pr edit 158 --repo luoyuctl/agenttrace --add-label lane/product --add-label status/ready-for-review
• git fetch origin pull/158/head:pr-158
• git worktree add ../agenttrace-review-pr-158 pr-158
• git diff --stat origin/master...HEAD
• git diff --check origin/master...HEAD
• git diff --name-status origin/master...HEAD
• git diff --unified=80 origin/master...HEAD -- internal/tui/tui.go internal/tui/tui_test.go internal/i18n/i18n.go
• rg -n "(?i)(secret|token|api[_-]?key|password|prompt|session_id|/Users/|platform|star|follow|money wasted)" internal/tui internal/i18n || true
• go test ./internal/tui -run 'TestList(DriverSummaryShowsDominantGroupsAndCounts|ViewShowsSelectedSessionSummary|ViewSelectedSummaryFitsNarrowCriticalList)|TestViewsRenderWithinTerminalWidth|TestViewsFitVerySmallTerminalBounds' -count=1
• go test ./...
• go build -o /tmp/agenttrace-review-pr-158 ./cmd/agenttrace
• /tmp/agenttrace-review-pr-158 --doctor || true
• /tmp/agenttrace-review-pr-158 --demo --overview -f json >/tmp/agenttrace-review-pr-158-demo.json
• wc -c /tmp/agenttrace-review-pr-158-demo.json

Decision:
PASS. This was a focused TUI usability improvement with local and CI validation passing. It was eligible for Quality auto-merge and has been squash-merged.

Label action: add status/auto-merged, remove status/ready-for-review.

Handoff to: Product
#144 is implemented for the critical-list scanability path.