[t3dotgg]

#22189 split a PR's +N −N into Code, Tests, Mocks, and Generated. There were three problems. The split only showed on the PR panel header. It waited for a GitHub read on first hover, so it felt slow. And Markdown and text changes counted as Code.

Now:

• PR panel: the split loads when the panel opens, so the hover shows it at once.
• Thread details card: the split shows inline under the files line. It loads when the card opens.
• PR list rows and a thread's linked PR rows: they get the same hover tooltip. Rows still load only on hover, so a long list does not spend a GitHub read per row.
• New Docs row: .md, .mdx, .markdown, .rst, .adoc, and .txt count as Docs. requirements*.txt, constraints*.txt, CMakeLists.txt, robots.txt, and llms.txt stay Code. Mobile's "Changed files" line shows Docs too.

The new docs key has a decoding default, so a newer client still reads a breakdown from a server that only sends four kinds.

Before (PR panel, #22117)	After (PR panel, #22197)	After (list row, #19426)

Verification

• Ran the dev build against real GitHub PRs. On fix(client): slow networks no longer drop working connections #19426, the row tooltip shows Code +61 −13, Tests +226 −36, Docs +12 −6. That matches gh pr view --json files. The PR panel request fires when the panel opens, before any hover.
• Contract test: a four-kind breakdown from an older server still decodes. Classifier and server breakdown tests cover Docs.
• New browser test: hovering the thread details row opens the card with the split. It fails without this change.
• Typecheck is clean for contracts, shared, server, web, and mobile. Lint is clean on the changed files. The ThreadDetailsPrRow browser tests pass.

Reviewed with sol-loop: 2 rounds with GPT-6.1 Sol. Sol verdict: approved

Created with Claude Opus 5.5 in Claude Code.

🤖 Generated with Claude Code

[macroscopeapp]

Approvability

Verdict: Needs Human Approval

Macroscope's review found this PR requires human approval — This PR adds a new Docs classification and exposes change-breakdown UI across PR panels, list rows, thread details, and mobile, including eager data loading in existing detail views. The cross-surface user-facing behavior and changed default request timing merit human review.

Notes:

• No code objects were reviewed. Approvability was decided on eligibility alone.

^{You can add or adjust custom eligibility rules. Learn more.}

[coderabbitai]

✨ Finishing Touches: 💡 1

🛠️ Fix failing CI checks: 💡

• Commit to this branch
• Create a new PR

📝 Generate docstrings

• Commit to this branch
• Create a new PR

🧪 Generate unit tests (beta)

• Commit to this branch
• Create a new PR

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

_{Comment @coderabbitai help to get the list of available commands.}

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	5.0 KiB	5.0 KiB	0 B (0.0%)	6.8 KiB	✅
Codex	Thread snapshot wire	3.8 KiB	3.8 KiB	0 B (0.0%)	4.9 KiB	✅
Codex	Live turn WebSocket wire	1.2 KiB	1.2 KiB	0 B (0.0%)	2.0 KiB	✅
Codex	Live turn WebSocket decoded	20.9 KiB	20.9 KiB	0 B (0.0%)	29.3 KiB	✅
Codex	Live turn messages	2	2	0 (0.0%)	8	✅
Claude	Total thread wire	5.0 KiB	5.0 KiB	−24 B (−0.5%)	6.8 KiB	✅
Claude	Thread snapshot wire	3.8 KiB	3.8 KiB	0 B (0.0%)	4.9 KiB	✅
Claude	Live turn WebSocket wire	1.2 KiB	1.2 KiB	−24 B (−2.0%)	2.0 KiB	✅
Claude	Live turn WebSocket decoded	21.2 KiB	21.2 KiB	−41 B (−0.2%)	29.3 KiB	✅
Claude	Live turn messages	2	1	−1 (−50.0%)	8	✅

Baseline: 774c6c1 · PR result: 44aac0f · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 108.5 KiB
• Claude decoded thread snapshot: 108.8 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.