[wSedlacek]

Reopened from closed upstream PR pingdotgg#31242 as a draft for t3.1code.
Idea rank: 12 of 108. Branch is as it was upstream; it likely needs a rebase (modernize) before review.

Problem

People who think in vim have to drop that habit the moment they write a prompt in T3 Code. There is no vim mode in the composer or any other text box, and moving around the app takes the mouse or a set of unrelated shortcuts.

Change

Adds an opt-in Settings → Chat → Vim mode for web and desktop. It is off by default, and the engine is lazy loaded, so nobody pays for it unless they turn it on.

• Editing: the composer and every multi-line text box get motions, operators, text objects, visual mode, ., registers, marks, macros, / search, and :s. Instead of writing our own vim, this drives the @​replit/codemirror-vim-core engine (exact pin, small local patch) through two adapters: one for ProseMirror and one for textareas. Chips and rich formatting survive yanks and pastes, and each insert session is one undo step.
• App navigation: outside a text box, j/k scroll, J/K change threads, : opens the command palette, g chords toggle panels, and i focuses the composer. Commands run through the user's configured shortcuts, so rebinding keeps working.
• Left alone: Enter still sends, app shortcuts keep working, and single-line fields and the terminal are unchanged. Mobile is not included: it has no hardware-keyboard-first editor to attach this to.

The full key list is in docs/user/keybindings.md.

There are no before images because nothing changes with the setting off.

Scope and approval

There is no linked issue or discussion, and no maintainer has approved this direction or scope. This is a new feature, so it does not qualify for an exemption. It is opened as a proposal, and I understand it may be closed under the contribution policy until that approval exists.

Verification

• vp run check:changed passes on the final commit, rebased on current main: format, lint, knip, typecheck, react compiler, and the tests beside the changes (2,442 passed).
• 323 key sequences with expected results generated by real Vim 9.1 run against the composer.
• The engine's own test suite runs against both the composer and textarea adapters. Cases that do not apply, such as % handling of code strings and comments, are listed by name in composer-vim.conformance.browser.test.ts.
• Composer browser tests use the real extensions and cover lists, task items, chips, undo order, and sending.
• In the real web app, I checked navigation, a textarea in a dialog, and composer editing. That pass ran before the last rebase.

After rebasing onto pingdotgg#30421, ** markers shown at the caret are editable text, so 0x on a bold word now removes a *. I updated the two affected tests to expect that. The rebase also needed one adapter change for the new splitOrLiftListItem argument. Both are covered by tests only, not by a new real-app pass.

Known differences from real Vim, found while checking review feedback: / and ? use JavaScript regex syntax (the engine's pcre default), so \+ is a literal plus. dit across lines does nothing. di( keeps no indented closing line, leaving () where Vim leaves (\n ).

Not tested: Electron, IME input, RTL text, Windows hardware, and in the real app the d, K, L, h, l, gb, gd, gf, gr keys and j/k in menus. Block visual mode is not supported.

Council review

Rounds: 7

Delegate	R1	R2	R3	R4	R5	R6	R7
Claude Opus	5/10	6.5/10	6/10	6/10	7/10	6.5/10	9/10
GPT Sol	6/10	7/10	7.5/10	8/10	8/10	8/10	9.5/10
Gemini	6.5/10	7.5/10	9.5/10	6.5/10	no score	9.5/10	8.5/10

The council reviewed the composer only. It did not review the textarea or navigation work, or the last two composer fixes made after round 7. Gemini only reviewed and wrote no code. The fixes for the bot review findings came after the council rounds too.

Created with Claude Opus 5.5 in Claude Code.

🤖 Generated with Claude Code

🤖 Generated with Claude Code

[github-actions]

This PR changes the test suite by +87 test cases (browser +87, unit 0, server 0).

New browser test files: apps/web/src/composer-vim.conformance.browser.test.ts, apps/web/src/composer-vim.reference.browser.test.ts, apps/web/src/vim-mode.browser.test.tsx.
Each browser test file costs CI seconds of module loading before its first test runs; a unit test costs milliseconds.

Keep tests that would fail for a real reason, and test logic below the UI where it can be. See Testing.

[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	21.5 KiB	21.5 KiB	0 B (0.0%)	24.4 KiB	✅
Codex	Live turn messages	1	1	0 (0.0%)	8	✅
Claude	Total thread wire	5.0 KiB	5.0 KiB	0 B (0.0%)	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	0 B (0.0%)	2.0 KiB	✅
Claude	Live turn WebSocket decoded	21.9 KiB	21.9 KiB	0 B (0.0%)	24.4 KiB	✅
Claude	Live turn messages	1	1	0 (0.0%)	8	✅

Other scenarios

All 64 scenario measurements match main.

All 64 scenario measurements

Provider	Scenario	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	short turn resume	WebSocket decoded	10.5 KiB	10.5 KiB	0 B (0.0%)	12.7 KiB	✅
Codex	short turn resume	WebSocket messages	2	2	0 (0.0%)	2	✅
Codex	long turn running, cold open	HTTP decoded	281.7 KiB	281.7 KiB	0 B (0.0%)	302.7 KiB	✅
Codex	long turn running, cold open	rows	75	75	0 (0.0%)	75	✅
Codex	long turn running, history page	HTTP decoded	92.4 KiB	92.4 KiB	0 B (0.0%)	107.4 KiB	✅
Codex	long turn running, history page	rows	75	75	0 (0.0%)	75	✅
Codex	long turn finished, cold open	HTTP decoded	381.5 KiB	381.5 KiB	0 B (0.0%)	410.2 KiB	✅
Codex	long turn finished, cold open	rows	151	151	0 (0.0%)	151	✅
Codex	long turn finished, socket cold open	WebSocket decoded	381.6 KiB	381.6 KiB	0 B (0.0%)	410.2 KiB	✅
Codex	long turn finished, socket cold open	WebSocket messages	2	2	0 (0.0%)	2	✅
Codex	long turn finished, socket cold open	rows	151	151	0 (0.0%)	151	✅
Codex	long turn finished, work page	HTTP decoded	186.2 KiB	186.2 KiB	0 B (0.0%)	200.2 KiB	✅
Codex	long turn finished, work page	rows	150	150	0 (0.0%)	150	✅
Codex	idle thread resume	WebSocket decoded	65 B	65 B	0 B (0.0%)	200 B	✅
Codex	idle thread resume	WebSocket messages	1	1	0 (0.0%)	1	✅
Codex	fleet shell snapshot	HTTP decoded	42.4 KiB	42.4 KiB	0 B (0.0%)	53.7 KiB	✅
Codex	fleet shell snapshot	rows	27	27	0 (0.0%)	27	✅
Codex	fleet shell subscribe	WebSocket decoded	65 B	65 B	0 B (0.0%)	200 B	✅
Codex	fleet shell subscribe	WebSocket messages	1	1	0 (0.0%)	1	✅
Codex	fleet shell resume check	HTTP decoded	18 B	18 B	0 B (0.0%)	100 B	✅
Codex	fleet shell resume check	rows	0	0	0	0	✅
Codex	stale shell reload	HTTP decoded	42.4 KiB	42.4 KiB	0 B (0.0%)	53.7 KiB	✅
Codex	stale shell reload	rows	27	27	0 (0.0%)	27	✅
Codex	stale shell resume	WebSocket decoded	140 B	140 B	0 B (0.0%)	400 B	✅
Codex	stale shell resume	WebSocket messages	2	2	0 (0.0%)	2	✅
Codex	stale todo list cold open	HTTP decoded	72.4 KiB	72.4 KiB	0 B (0.0%)	83.0 KiB	✅
Codex	stale todo list cold open	rows	75	75	0 (0.0%)	75	✅
Codex	stale todo list cold open	nodes	2	2	0 (0.0%)	4	✅
Codex	stale todo list cold open	subagents	2	2	0 (0.0%)	4	✅
Codex	stale todo list cold open	plans	1	1	0 (0.0%)	2	✅
Codex	stale todo list cold open	providerTurns	1	1	0 (0.0%)	2	✅
Codex	stale todo list cold open	checkpoints	1	1	0 (0.0%)	2	✅
Claude	short turn resume	WebSocket decoded	10.6 KiB	10.6 KiB	0 B (0.0%)	12.7 KiB	✅
Claude	short turn resume	WebSocket messages	2	2	0 (0.0%)	2	✅
Claude	long turn running, cold open	HTTP decoded	283.3 KiB	283.3 KiB	0 B (0.0%)	302.7 KiB	✅
Claude	long turn running, cold open	rows	75	75	0 (0.0%)	75	✅
Claude	long turn running, history page	HTTP decoded	93.4 KiB	93.4 KiB	0 B (0.0%)	107.4 KiB	✅
Claude	long turn running, history page	rows	75	75	0 (0.0%)	75	✅
Claude	long turn finished, cold open	HTTP decoded	384.2 KiB	384.2 KiB	0 B (0.0%)	410.2 KiB	✅
Claude	long turn finished, cold open	rows	151	151	0 (0.0%)	151	✅
Claude	long turn finished, socket cold open	WebSocket decoded	384.3 KiB	384.3 KiB	0 B (0.0%)	410.2 KiB	✅
Claude	long turn finished, socket cold open	WebSocket messages	2	2	0 (0.0%)	2	✅
Claude	long turn finished, socket cold open	rows	151	151	0 (0.0%)	151	✅
Claude	long turn finished, work page	HTTP decoded	188.1 KiB	188.1 KiB	0 B (0.0%)	200.2 KiB	✅
Claude	long turn finished, work page	rows	150	150	0 (0.0%)	150	✅
Claude	idle thread resume	WebSocket decoded	66 B	66 B	0 B (0.0%)	200 B	✅
Claude	idle thread resume	WebSocket messages	1	1	0 (0.0%)	1	✅
Claude	fleet shell snapshot	HTTP decoded	43.0 KiB	43.0 KiB	0 B (0.0%)	53.7 KiB	✅
Claude	fleet shell snapshot	rows	27	27	0 (0.0%)	27	✅
Claude	fleet shell subscribe	WebSocket decoded	66 B	66 B	0 B (0.0%)	200 B	✅
Claude	fleet shell subscribe	WebSocket messages	1	1	0 (0.0%)	1	✅
Claude	fleet shell resume check	HTTP decoded	18 B	18 B	0 B (0.0%)	100 B	✅
Claude	fleet shell resume check	rows	0	0	0	0	✅
Claude	stale shell reload	HTTP decoded	43.0 KiB	43.0 KiB	0 B (0.0%)	53.7 KiB	✅
Claude	stale shell reload	rows	27	27	0 (0.0%)	27	✅
Claude	stale shell resume	WebSocket decoded	140 B	140 B	0 B (0.0%)	400 B	✅
Claude	stale shell resume	WebSocket messages	2	2	0 (0.0%)	2	✅
Claude	stale todo list cold open	HTTP decoded	73.4 KiB	73.4 KiB	0 B (0.0%)	83.0 KiB	✅
Claude	stale todo list cold open	rows	75	75	0 (0.0%)	75	✅
Claude	stale todo list cold open	nodes	2	2	0 (0.0%)	4	✅
Claude	stale todo list cold open	subagents	2	2	0 (0.0%)	4	✅
Claude	stale todo list cold open	plans	1	1	0 (0.0%)	2	✅
Claude	stale todo list cold open	providerTurns	1	1	0 (0.0%)	2	✅
Claude	stale todo list cold open	checkpoints	1	1	0 (0.0%)	2	✅

Baseline: 7fd5a75 · PR result: 2560c25 · 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.