Round two of adversarial review. The skeptic confirmed the session_activity finding
with extra evidence from this host and raised three things the first pass missed.
SECURITY — `tmux -t <name>` resolves exact name, then NAME PREFIX, then fnmatch.
The whole module rests on "we only ever touch `web_` + a UUID v4, so a tmux session
the user made for their own work is never enumerated and never offered for
deletion". Prefix matching goes around that: a session named `web_<uuid>_mine` is
refused by parseSessionList (its suffix is not a UUID) and so never appears in the
listing — but `-t web_<uuid>` matches it. So a DELETE aimed at an id that does not
exist would end the user's session, and a preview would print its screen.
Verified on tmux 3.6a, isolated socket:
has-session -t web_<uuid> → succeeds against web_<uuid>_MINE
has-session -t =web_<uuid> → "can't find session"
capture-pane -p -t web_<uuid> → printed the DECOY's screen
capture-pane -p -t =web_<uuid>: → "can't find session"
Two target forms are needed, because `=` only qualifies the session part of a
target: session targets (has-session, kill-session) take `=name`, while
capture-pane takes a PANE and rejects a bare `=name` outright ("can't find pane"),
so it needs `=name:`. Both are now the only way a name reaches tmux, with an
integration test that stands up a real decoy and asserts the listing omits it and
both the preview and the DELETE report 404 while it stays alive.
CORRECTNESS — take max(window_activity, session_activity), not window_activity
alone. The previous commit swapped one clock for the other, but they are not
ordered: window_activity tracks pane output while an attach with no output bumps
only session_activity. On this host one session's session_activity was 4 s NEWER
than its window_activity (and another's window_activity was 25 days newer). The
question is "has anything happened here", so the answer is the later of the two.
Also documents the caveat that window_activity resolves against the session's
current window only — exact for the single-window sessions this app creates, and
bounded by the max for anything else.
CORRECTNESS — parseSessionList accepted an empty numeric field. Number('') is 0
and 0 is finite, so a blank clock parsed as "created at the epoch, idle ever
since": instantly eligible for the idle cleanup. Fields must now be actual digits.
SAFETY — bulk cleanup re-reads the world immediately before killing. The candidate
list is a snapshot and the kill loop takes time, so a session could be attached, or
adopted into the table, inside that window. This is the irreversible path; it now
verifies each candidate against a fresh listing rather than trusting the snapshot.
PERFORMANCE — captureOrphan no longer probes with hasSession first. That was a
second tmux spawn per thumbnail, and a SYNCHRONOUS one on the path a whole grid
refreshes, buying nothing: capture-pane already fails cleanly on a missing session
and we already turn that into null. The guard splits into a pure half
(mayActOnOrphan: UUID shape + not in the table) used by both paths, and the
existence probe, which only the destructive path needs.
Tests: 2213 unit, 9 orphan integration (incl. the decoy regression). Confirmed
`has-session -t =<name>` still resolves the 69 real sessions, so the Case 3.5
re-attach path is unaffected.
Note: test/integration/server.test.ts "H1 shell state survives a server restart"
flakes ~1 run in 3 when all 27 run together on a loaded machine, and passes 4/4 in
isolation. That is the pre-existing real-PTY timing flake already recorded in
PROGRESS_LOG, not this change — it exercises Case 3.5, which is verified above.
Adversarial review of the previous commit found ten defects. The critical one
would have destroyed exactly the work this feature exists to protect.
CRITICAL — `#{session_activity}` is a CLIENT clock, not an output clock.
It advances on attach and on keypresses, and never moves for a detached session
however much its shell is printing. Measured on tmux 3.6a: a detached session
running `while true; do echo tick; sleep 1; done` still reported
session_activity == session_created after 6 s, while `#{window_activity}` tracked
the output exactly. So "idle for 7 days" was really "nobody has typed for 7 days",
and `Clean up idle 7d+` would have killed long-running unattended sessions that
were busy working — the walk-away case the whole app is built around.
Measured against this host's 69 sessions: the old clock condemned 36, the new one
condemns 33, so three sessions doing real work were one click from deletion.
Now reads window_activity, with a test that fails if the format string regresses.
HIGH — no timeout on any tmux call. maxBuffer bounds a flood but nothing bounded
the wait, and a synchronous exec cannot be interrupted by the event loop, so a
tmux server stopped mid-syscall hung the whole server permanently. All five calls
(including the three that predate this feature) now carry one.
HIGH — listSessions and capturePane were synchronous on a POLLED path. Measured
73 ms and 58 ms per call on this host; every home screen on every device refreshes
on a 5 s timer, and each of those was 73 ms with the event loop stopped, i.e.
73 ms in which no PTY byte reached any terminal anywhere. Both are async now, so
the manager's listOrphans/captureOrphan/killOrphansIdleSince are too. The server
is a byte-shuttle first: nothing polled may block it.
HIGH — GET /orphan-sessions had no rate limiter although, unlike every other read
route, it spawns a subprocess. It shares the preview bucket now.
HIGH — the cleanup confirmation counted CARDS, and the grid is capped at 24 while
the cleanup matches across every session the server enumerates. On this host it
would have said "24" and ended 33 shells. New GET /orphan-sessions/count-idle
answers the real question, and the dialog states that number, or refuses to
proceed if it cannot be determined.
MEDIUM — fetchOrphanSessions collapsed every failure into [], so one 429 or 500
read as "all recovered sessions are gone", tearing down every mounted xterm and
rebuilding it on the next success. It returns null for failure now, and the
refresh leaves the section untouched.
MEDIUM — the preview latch was "card was created", not "preview succeeded", so a
card that lost its single request stayed blank forever. Latches on success.
MEDIUM — overlapping refreshes. refresh() is entered from a 5 s timer, from
killOrphan and from cleanup, with awaits inside and no guard, so a slow poll could
resurrect cards for sessions already killed or re-mount them into a root that
setVisible(false) had just cleared. A generation counter, bumped on teardown and
checked after every await, invalidates stale runs.
MEDIUM — renderPreview could write to a disposed xterm (it throws "Object has
been disposed"). PreviewCard carries an explicit `disposed` flag set by a new
disposeCard() helper, so no caller can dispose a terminal and forget the flag.
Deliberately NOT el.isConnected: a card not yet appended is alive, and two
existing preview-grid tests correctly caught that conflation.
LOW — kill and cleanup discarded their results, so a refusal (403 behind a proxy
whose Origin differs, 404 on a lost race) looked like a dead button. Both report.
Tests: 2209 unit, 27 e2e, 8 orphan integration. Verified live against all 69 real
sessions with none harmed; the window_activity fix confirmed on an isolated tmux
socket (-L) so the developer's sessions were never involved.
A `web_*` tmux session that outlives the process which created it was unreachable
from the UI, permanently. Nothing enumerated tmux — src/session/tmux.ts had
hasSession but no list — so recovery only worked if a client still had the id in
localStorage. list() walks the in-memory table, and so does reapIdle, so such a
session could not be listed, previewed, joined or killed, and never aged out.
On this host that was 69 tmux sessions with 5 clients: 64 unreachable, the oldest
from 26 Jun, one of them still running a Claude Code session with 7 agents.
This is the "catalogue" approach: make them visible and actionable WITHOUT
adopting them. Nothing is attached, because attaching makes tmux resize the window
to the new client's dimensions and SIGWINCH whatever is running inside — doing
that to dozens of live TUIs at startup is exactly the outcome to avoid. Adoption
stays an explicit user action: opening an orphan re-attaches by id through the
existing re-attach path, and it becomes an ordinary live session.
src/session/tmux.ts listSessions() + parseSessionList() (pure, so the filter
rules are unit-testable) and capturePane(), which reads a
session's current screen via `capture-pane -p -e` — no
client, no resize, colour preserved.
src/session/manager.ts listOrphans/captureOrphan/killOrphan/killOrphansIdleSince
src/server.ts GET /orphan-sessions, GET /orphan-sessions/:id/preview,
DELETE /orphan-sessions/:id, DELETE /orphan-sessions
?idleDays=N
public/ a "Recoverable" section under the home grid, reusing the
existing card/thumbnail machinery via orphanAsLive()
Guards, all enforced in the manager so no route can forget one:
- the tmux name must be `web_` + a UUID v4 (SESSION_ID_RE, the same gate the
attach protocol applies). This is what stops a hostile or malformed name from
reaching `tmux -t` — a literal `-t`, or path traversal — and it means a tmux
session the user created for their own work is never enumerated, never
previewed and never offered for deletion.
- an id the table already owns is refused everywhere: killing it via tmux would
end the shell behind a live PTY's back and leave a zombie table entry. Those go
through killById.
- every function is inert when cfg.useTmux is off, so the non-tmux deployment is
byte-for-byte unchanged.
- bulk cleanup skips ATTACHED sessions and requires idleDays >= 1: there is no
"delete everything" form of the route. "This server does not track it" is not
evidence that it is abandoned — a plain `tmux attach` and a second server
process both report as attached.
- reads carry no Origin guard (same threat model as /live-sessions); both DELETEs
do. Preview gets its own rate-limit bucket since each call spawns a process.
Two things the live run exposed and this fixes: previews were loading
sequentially (69 tmux spawns per 5 s refresh) — now once per card, fire-and-forget
— and the grid is capped at 24 cards because each thumbnail is an xterm instance.
The remainder is stated in the UI with the tmux command to reach it, never
silently truncated.
Separate wire type from LiveSessionInfo on purpose: an orphan supports none of the
live-session operations, and the Android/iOS clients decode /live-sessions, so a
variant shape there would break them.
Tests: 2203 unit (+42: parse/filter rules, all four manager guards, tmux-off
inertness) and 8 integration against real tmux — create a throwaway session,
list it, capture it while asserting it stays unattached, reject non-UUID ids,
reject a foreign Origin, kill it, 404 the second time. The integration file never
issues the bulk DELETE: it runs against the host's real tmux server and would end
the developer's own sessions. Verified live against all 69 with none harmed.
The low-severity leftovers from the panel audit, plus the tooling trap that made
reviewing this work harder than it needed to be.
.gitattributes (new) — `public/projects.ts diff`
That file embeds literal NUL bytes on purpose: the namespace sentinels
'\0active' / '\0other' are prefixed so they cannot collide with a real
`First.Second` key. git therefore classified it as binary and printed
"Binary files differ" for every diff, and `git diff --stat` reported
"Bin 66815 -> 68701 bytes" instead of line counts — the file was effectively
unreviewable. The `diff` attribute makes git diff it as text without setting
`text`, so no line endings are normalised and the bytes on disk are untouched.
(Same trap in the shell: grep prints only "Binary file ... matches", which
reads as a no-match. The comment in the file says to use grep -a.)
Detached HEAD now shows its short sha
The Head cell held the "detached HEAD" chip and nothing else: it said what was
wrong and nothing about where you were. SyncState gains an optional `head`,
populated from the .git/HEAD text readSyncState ALREADY reads for the detached
test — so this costs no extra I/O and no extra git call. Guarded by a hex regex
and rendered as textContent.
One geometry for the ghost button role
The design defines two button roles. The outline one had drifted to three
different sizes (.proj-wt-open at 11.5px/4px 12px/6px, .proj-diff-toggle at
12.5px/7px 15px/7px) and both used full-strength --text instead of the muted
label. All four outline buttons are now declared once, together, at the
design's geometry; the duplicate .proj-diff-toggle block further down the file
is gone, since a second block at equal specificity is exactly how the role
drifted in the first place.
Header branch chip joins the chip language
.proj-branch was the only chip with no border and no mono face, and 1px tighter
all round, so the detail header's branch did not match the branch chips in the
worktree list right below it.
"Changes" heading over the View Diff toggle
Every other block on the page is introduced by a section heading. Without one
the toggle hung 4px off the bottom of the sync band and read as part of it.
Tests: 2167 unit (4 new — detached sha present/absent on both the server and the
panel) and 27 integration, twice.
The page carried one bare `●` for "dirty" and nothing else, so "do I have
commits I haven't pushed" and "which worktree am I in" still meant dropping
into a terminal. Design mock: docs/mockups/project-detail-git.html; plan and
task breakdown (G1-G7): docs/plans/w6-project-git-panel.md.
One rule drives the whole feature: ahead/behind compare against `@{u}`, a
LOCALLY CACHED remote ref that only a fetch moves. This repo was the live
example while building — `↑9` true, `↓0` false, because FETCH_HEAD had not
moved in 19 days. So:
- `ahead` needs only local refs and is never flagged.
- `behind` is flagged `stale` once FETCH_HEAD is older than an hour.
- Exactly ONE state may render green: ↑0 ↓0 AND a fresh fetch. Green means
"I checked, ignore this"; getting it wrong is lying to the user.
- No upstream (the normal state of a fresh worktree branch) leaves ahead and
behind undefined — it renders an explicit `no upstream`, never the green
path. That fall-through is the easiest bug to ship here.
What landed:
G1 SyncState (upstream/ahead/behind/lastFetchMs/detached) + ProjectDetail.sync
and .dirtyCount. All additive and optional — the Android and iOS clients
decode these shapes. The ahead/behind helper already existed for the list
view; buildProjectDetail had simply never called it.
Fixes a pre-existing bug on the way: readBranch read <repo>/.git/HEAD
directly, so it returned nothing inside a LINKED worktree, where .git is a
file. resolveGitDirs now resolves both the per-worktree gitdir (HEAD) and
the shared common dir (FETCH_HEAD).
G2 POST /projects/git/fetch. Same discipline as push: the remote is derived
server-side and no remote or refspec is ever read from the body, so a
client cannot aim it at an arbitrary URL. Touches refs/remotes only — no
working tree, no index, no merge; it is not a pull. Own rate-limit bucket
so refreshes cannot eat the budget a real push needs. On failure
lastFetchMs is left alone, so the UI keeps saying "stale" instead of
pretending it refreshed.
G3 makeSyncBand replaces the bare dot: upstream name, ↑n, ↓n, stale flag,
dirty count, Fetch button (disabled on a detached HEAD).
G4 The commit list marks unpushed commits and draws the upstream boundary
once, after the last of them. Marking is server-side from `rev-list`,
deliberately NOT "the first N rows": `git log` is date-ordered, so merging
an older branch interleaves unpushed commits BELOW pushed ones, and that
shortcut fails in the dangerous direction — calling an unpushed commit
pushed. A regression test builds exactly that backdated-merge shape.
G5 The worktree section is always "Worktrees (n)" (it used to rename itself
to "Branch" at n=1) and the current row carries its own state chips.
G6 Cost control. The plan called for a .git-mtime cache; that was dropped
during implementation because a fingerprint over HEAD/index/reflog does
NOT move when push updates a remote-tracking ref — the cached `ahead`
would still claim "9 to push" right after a successful push, which is the
exact lie the feature exists to prevent. Replaced with three measures that
cannot go stale: in-flight coalescing (N devices watching one repo cost
one probe, entry dropped as it settles, nothing cached across time),
skipping the re-render when the payload is byte-identical (this also stops
the 5 s re-mount of the commit log, two more git spawns per tick), and
pausing the timer while the document is hidden.
G7 Per-worktree state via GET /projects/worktree/state, kept narrower than
/projects/detail so N rows do not pay for worktree listing and CLAUDE.md
reads nothing renders. Needed an unplanned prerequisite: ProjectSessionRef
carried no cwd, so sessions could not be attributed to a worktree. Added
it, plus countSessionsByWorktree, which matches DEEPEST-first because
.claude/worktrees/<name> lives INSIDE the main checkout and prefix
matching would count every worktree session against the parent repo too.
Out of scope, unchanged: no reset, no checkout, no clean, no rebase, no
force-push. stage/commit/push stay exactly as they were.
Verified: tsc and build clean; 46 new tests.
A launchd/Finder-launched server has no LANG/LC_* at all, so tmux — which
decides per client whether the terminal is UTF-8 capable purely from those
variables — fell back to its non-UTF-8 mode and rewrote the byte stream
server-side, before it ever reached xterm:
- every wide character became `_` (中文测试 → ________)
- `✓` / `⏺` / emoji became `_`
- `═║╔╗` was downgraded to DEC Special Graphics (ESC ( 0), so rounded
corners vanished and boxes rendered as loose horizontal lines
The `-l` login shell cannot fix this: tmux is the PTY's root process and has
already made its decision by the time the shell sources ~/.zprofile.
withUtf8Locale() fills a locale in on the spawn env, and only when needed —
an existing UTF-8 LC_ALL / LC_CTYPE / LANG (zh_CN.UTF-8, ja_JP.UTF-8, …) is
left untouched; only a missing or non-UTF-8 charset is replaced.
Existing tmux sessions do not need recreating: re-attaching with a UTF-8
client restores correct output (already-scrolled `_` stays mangled).
An OPTIONAL shared token so the app can be used off-LAN (via the relay/tunnel) more
safely than "anyone who reaches the port gets a shell". Sits IN FRONT OF the existing
Origin/CSRF model — never replacing it — and is fully inert when unset.
- src/http/auth.ts (new, pure): constantTimeEqual hashes both inputs to sha256 (32
bytes) then crypto.timingSafeEqual — no `===`, no length side-channel, never throws.
parseCookieHeader / cookieIsAuthed / buildSetCookie / isAuthEnabled.
- WEBTERM_TOKEN in config: 16–512 URL/cookie-safe chars or the server refuses to start.
- GET /?token=<t> or POST /auth (rate-limited 10/min) validates → sets
HttpOnly; SameSite=Strict; Secure-when-https cookie; public/login.html (no <script>).
- When enabled: the WS handshake (AFTER the Origin check, before handleUpgrade) + a
central authGate over all remote HTTP require the cookie. Open: /login, /auth.
Loopback bypass scoped to /hook* ONLY (tighter than the plan — gates the local
browser too).
- Unset ⇒ isAuthEnabled false ⇒ pure passthrough (LAN zero-config unchanged).
Honest tradeoff (in code + CLAUDE.md + login page): a bar-raiser, NOT a TLS/Tailscale
substitute — on bare ws:// the token is cleartext + replayable; only hardens the
TLS-terminated relay path. Verified: typecheck + build:web clean, 2118 pass at
--test-timeout=30000 (disabled-mode regression proven; only the known tmux flake red);
auth.ts 100% line coverage.
Fan ONE prompt across N branch/agent lanes: N worktrees, N Claude sessions on the
same prompt, watched side-by-side in the split-grid board, approve/kill per lane,
🏆 keep the winner (losers' worktrees removed). ~90% composition of shipped parts.
- src/http/session-groups.ts (new, pure, no git exec): deriveRepoRoot /
groupSessionsByRepo (cluster sessions by their <repo>-worktrees parent).
- GET /live-sessions/grouped (read-only; registered BEFORE /live-sessions/:id so
"grouped" isn't captured as an id). MAX_FANOUT_LANES env (default 6, = grid-6 cap).
- public/fanout.ts (new, pure): buildFanoutCmd shell-quotes the prompt (single-quote
wrap with '\''-escaping so $(...)/backticks/;/&& can't execute) + collapses newlines
+ caps 4000 chars; laneBranch/slugify. Effective N = min(lanes, maxFanoutLanes, 6),
≥2; the maxSessions cap is enforced server-side ("Started K of N" banner).
- public/tabs.ts launchFanout (N× createWorktree → openProject w/ prompt pre-injected
via the existing initialInput) + keepFanoutWinner; public/projects.ts renderFanoutForm
+ extracted shared createWorktreeReq (DRY). Byte-shuttle preserved (lane = own PTY).
One-click merge deferred (winner session stays open for a manual merge).
Reused unchanged: createWorktree/removeWorktree, addEntry/initialInput, split-grid +
per-quadrant approve/maximize/monitor + gauges. Verified: typecheck + build:web clean,
2063 pass at --test-timeout=30000 (only the known tmux/PTY flake red). Fixed 3
/config/ui exact-shape tests to include the new maxFanoutLanes field.
The walk-away endgame — review a diff on your phone, then land it without typing
git into a mobile terminal. MVP bounded to per-file stage/unstage, commit, and
push the current branch; discard/checkout/reset deliberately deferred (nothing here
mutates working-tree file contents).
- src/http/git-ops.ts (new): stageFiles/commit/push (execFile, no shell, never
throws). Path containment: every files[] entry realpath-contained under the repo,
argv after `--`, capped at diffMaxFiles. Commit message: empty/over-5000 → 400,
single -m argv. Push: current branch only — has-upstream → plain `git push`;
no-upstream + one remote → `git push -u <remote> <branch>`; 0 remotes → 400,
≥2 → 409, detached HEAD → 400. Remote AND branch read from the repo, never the
client. NEVER --force / +refspec. GIT_TERMINAL_PROMPT=0 + ssh BatchMode fail auth
fast (401). Errors classified to fixed safe strings (raw stderr never surfaced).
- POST /projects/git/{stage,commit,push} — all behind requireAllowedOrigin +
GIT_OPS_ENABLED kill-switch + per-IP rate limits (stage/commit 30/min, push 6/min)
+ isValidGitDir. public/diff.ts: per-file Stage/Unstage + a commit/push bar
(all text via textContent; re-loads the diff on success).
Verified: typecheck + build:web clean, git-ops tests 213 pass (unit covers escape
rejection / empty+overlong commit / push argv upstream-vs-no-upstream / classified
errors), full suite green at --test-timeout=30000.
Closes the create-only loop — delete losing worktrees and prune stale ones without
a terminal. Destructive, so guarded hard:
- src/http/worktrees.ts: removeWorktree + pruneWorktrees (execFile, no shell, never
rm -rf; git worktree remove [--force] / git worktree prune). Safeguards:
(1) target realpath must match an entry git itself reports in `git worktree list`
for THIS repo → 404 otherwise (arbitrary FS paths never match, so git is never
invoked against them); (2) reject the MAIN worktree → 400; (3) realpath
containment on request + each list entry, operating on git's canonical path (+ --);
(4) dirty tree without force → 409 "force required"; (5) locked → 409; (6) errors
classified to fixed safe strings (raw stderr/paths never leaked).
- DELETE /projects/worktree + POST /projects/worktree/prune — both behind
requireAllowedOrigin + worktreeEnabled + audit-logged, mirroring the create route.
- public/projects.ts: ✕ remove button per linked-worktree row (hidden for main/locked)
+ a prune button; force needs an explicit second confirm (no single-click data loss).
Verified: typecheck + build:web clean, worktree tests 108 pass (unit covers
reject-non-registered / reject-main / reject-escape / refuse-dirty-without-force /
clean-remove / force-remove / prune), full suite green at --test-timeout=30000.
Four small, high-delight features that turn passive capture into glanceable signals.
- Sync chip on project cards: ahead/behind vs upstream + last-commit time, folded
into the existing concurrent per-repo metadata pass (git rev-list --count
--left-right @{u}...HEAD + git log -1 --format=%ct; no upstream → undefined, no route).
- Cost budget guard: COST_BUDGET_USD env (0 = off); a per-session one-shot latch
(Session.budgetNotified, cost is monotonic so never re-armed) fires a single push
on threshold crossing in manager.handleStatusLine; the already-broadcast telemetry
frame carries the warn (tg-cost-warn styling derived from costUsd>=budget via
/config/ui — no new ServerMessage). web-push title added to sw-push.js.
- "While you were away" digest: GET /digest?since= → {finished, needsInput, stuck,
totalCostUsd, sessions[]} aggregate over manager.list(); FE banner on reconnect.
- Recent commits per project: src/http/git-log.ts (NUL-delimited git log → CommitInfo[]),
GET /projects/log?path= (isValidGitDir), textContent-inert render in project detail.
All git via execFile (no shell) + validated cwd; new routes read-only; commit
messages rendered via textContent. Verified: typecheck + build:web clean, 1904 pass
at --test-timeout=30000 (two default-5s failures are slow-sandbox real-subprocess
timeout flakes — the known ring-buffer test + a new real-git-clone sync test — not
logic regressions).
Per-project chip: PR state · N checks passing · mergeable — glance from the phone,
re-engage only when it's red, instead of dropping into a terminal for `gh pr checks`.
- src/http/gh.ts (new): single `gh pr view --json number,state,title,url,isDraft,
mergeable,headRefName,baseRefName,statusCheckRollup` (execFile, no shell, cwd =
isValidGitDir repo, timeout + maxBuffer). summarizeChecks rolls the mixed
CheckRun/StatusContext rollup into {total,passing,failing,pending}. Never throws —
degrades to not-installed (ENOENT) / unauthenticated / no-pr / error / disabled.
Cache keyed by repoPath+branch (reuses projectScanTtlMs) with in-flight dedupe.
- src/types.ts: additive PrStatus/PrAvailability/PrCheckSummary; config GH_ENABLED
(default on) + GH_TIMEOUT_MS (8s).
- GET /projects/pr?path= (read-only, isValidGitDir); public/gh-chip.ts render-only
chip mounted in the project detail header (git repos only).
Read-only, host's own authed gh (same trust as the shell). No untrusted argv (only
the validated cwd); gh stdout/token never logged; the attacker-controllable PR title
is rendered inert via textContent (SEC-H4). Verified: typecheck + build:web clean,
1816 pass (gh tests 118). The 1 red is the known real-PTY ring-buffer timeout flake.
The git-diff viewer can now diff the current branch against a base commit-ish
(e.g. main) — review an agent's whole branch from your phone before merging, not
just uncommitted changes. Completes the long-deferred FR-B1.9.
- src/http/diff.ts: getDiff() gains an optional base. Three-layer defense so an
attacker-supplied base never reaches a shell or acts as a git option:
(1) isPlausibleRev() rejects leading '-', '..'/'...' ranges, metachars, control
chars, >250 chars; (2) git rev-parse --verify --quiet --end-of-options
<base>^{commit} — only a resolved 7-64 hex sha is accepted, else empty result;
(3) git diff --no-color <sha>... -- (three-dot = the branch's changes since
divergence, PR-style). execFile, no shell, timeout + maxBuffer bound.
- src/types.ts: additive optional base on DiffResult.
- src/server.ts GET /projects/diff reads ?base (400 on !isPlausibleRev); read-only.
- public/diff.ts: a "compare base" <select> (Working tree + one option per base),
disables Working/Staged tabs in base mode. public/projects.ts derives bases from
the worktree branches ∪ current branch.
Verified: typecheck + build:web clean, 1763 pass (diff tests 87 green). The 1 red
in the plain full run is the pre-existing real-PTY "ring buffer" flake (times out
at default 5s under sandbox load; 27/27 at --test-timeout=30000) — unrelated.
The walk-away primitive: queue follow-up prompts and let a session advance itself
while you're gone — "run the tests" drains, and when Claude next goes idle "open a
PR" drains. Injection reuses writeInput (byte-identical to a keystroke, broadcasts
to all mirrored devices); the byte-shuttle is untouched.
- POST/GET/DELETE /live-sessions/:id/queue — POST/DELETE behind requireAllowedOrigin
+ a per-IP rate limit (QUEUE_RATE_MAX=20/min) + SESSION_ID_RE; text non-empty,
byte-capped (QUEUE_ITEM_MAX_BYTES=4096, 16kb body), count-capped (QUEUE_MAX_ITEMS=10).
- Bounded per-session queue in manager (enqueueFollowup/drainOne/clearQueue);
new optional queueLength on LiveSessionInfo.
- Idle drain: the Stop/SessionEnd /hook branch scheduleDrain()s a debounced timer
(QUEUE_SETTLE_MS=1500). It drains exactly one entry only if the session still
exists, hasn't exited, produced no output during settle, and is idle — three
guards (debounced single timer + pop-one + settle re-check) → one entry per idle.
- New additive ServerMessage {type:'queue',length} broadcasts the count to mirrors
(⧗N badge); FE enqueue via the quick-reply editor + TabApp.enqueueToActive.
Injected bytes go verbatim to the PTY (never built into a shell command). Verified
independently: typecheck + build:web clean, 1737 tests pass (real-PTY integration
covers idle-drain / one-per-idle / settle-guard). Foundation for templated launches,
auto-continue, and issue-intake (docs/ROADMAP.md).
Remote one-tap approval was blind (you'd tap Approve without seeing Claude wants
to run `rm -rf` or rewrite a config). The pending tool's actual command / diff now
renders above the approval bar, on every attached device, riding the same broadcast
+ late-joiner rails as the existing `gate` field.
- src/http/approval-preview.ts (new, pure, never-throws): deriveApprovalPreview —
Bash → command; Edit/Write/MultiEdit/NotebookEdit → a synthetic DiffFile; else null.
Every line sanitized via sanitizeField (strips control/ANSI); caps 40 lines /
200 chars/line / 4KB (security limits, UTF-8-safe byte clamp).
- src/types.ts: additive optional `preview?: ApprovalPreview` on the status
ServerMessage + handleHookEvent (older clients ignore it).
- src/server.ts /hook/permission: derive preview from tool_input, store on the
PendingApproval entry, re-send to late joiners exactly like `gate`.
- manager.ts threads it; public/tabs.ts renderApprovalPreview (command → <pre>
textContent; diff → reused innerHTML-free renderDiffFile). Unknown tools /
plan gates fall back to today's name-only bar.
Attacker-influenced tool input → rendered via textContent/diff-renderer only,
never innerHTML. Verified independently: typecheck + build:web clean, 1692 pass.
Terminal output URLs and file paths (e.g. src/app.ts:42) are now tappable — the
walk-away device is a phone, so this removes soft-keyboard copy gymnastics.
- public/link-paths.ts (new, pure): findPathMatches — links tokens with a '/',
a :line suffix, or a code extension (src/app.ts:42, /abs/main.rs:10, README.md)
while rejecting example.com / v1.2.3 / 12:34 / URL tails.
- terminal-session.ts: hardened URL handler (scheme allowlist http/https/mailto +
window.open noopener,noreferrer, blocks javascript:/data:/file:) replacing the
addon default; a path link provider → openPath() POSTs {file,line} to
/open-in-editor (resolves rel paths against the OSC-7 cwd; in-flight guard).
- src/http/editor.ts: additive openFileInEditor + isGotoEditor (--goto file:line
only for goto-capable editors; execFile, no shell). openInEditor untouched.
- src/server.ts: /open-in-editor branches on body.file vs body.path (same CSRF guard).
Deviation from the "no server change" brief: an additive openFileInEditor was
required because the existing route only opens directories, not file:line (Option A
in docs/plans/w1-clickable-links.md). Backward-compatible; src/types.ts untouched.
Verified independently: typecheck + build:web clean, 1661 tests pass (link-paths
95%+ cov). Note: xterm buffer-row indexing (getLine(n-1)/range.y=n) is asserted in
jsdom but merits a real-browser smoke check.
Mirror src/push/apns.ts behind the same NotifyService seam (combineNotifyServices),
so held-gate/done hook events fan out to Android via data-only high-priority FCM with
zero new event paths. OAuth2 bearer via google-auth-library (R7); payload minimized to
sessionId/cls/token (never cwd/command/bytes); intentionally-loose FCM-token validator;
Origin-guarded + rate-limited POST/DELETE /push/fcm-token; FCM_* env all-or-disabled.
Cross-review fix: dropped validateStatus:()=>true so google-auth-library's built-in
401 refresh-and-retry actually fires. 60 new tests; full server suite green; tsc clean.
Spawn the shell with -l (POSIX only) so it loads the full profile chain
and rebuilds PATH. Without it, a shell spawned by a GUI-launched app or a
long-lived tmux keepalive server inherits a minimal PATH, so nvm-managed
node (only on the login PATH) is missing and Claude Code hooks fail with
'node: command not found'. PowerShell (Windows) has no -l; skipped there.
Applies to both the tmux new-session command and the direct-spawn path.
The projects panel listed bare parent folders (~, ~/Documents) as projects and
a single session lit up every ancestor project's 'Active now'. dropParentFolders
filters non-git entries that merely contain other listed projects; assignSessions
attributes each live session to the DEEPEST containing project (detail page keeps
prefix matching). +3 tests; projects 29/29, tsc clean. Frontend unchanged.
The flat Projects grid was overwhelming at 50-100 repos. Explored the UX with
4 parallel agents (auto-grouping / manual folders / density-segmentation /
competitive research); all converged on namespace auto-grouping as the primary
fix — the dotted repo names (Billo.Platform.*, Billo.Infrastructure.*) already
encode the hierarchy, so it's a string-split, not a taxonomy the user builds.
Frontend:
- groupProjects(): depth-2 namespace grouping, >=2-member threshold (singletons
fall to "Other"), degrade-to-flat when repos share no prefix, running projects
duplicated into a pinned "Active now" section.
- displayLabel(): cards show the namespace tail (Payment, not
Billo.Platform.Payment); full name in Active/Other.
- Collapsible sticky group headers with an "N active" badge so a collapsed
section never hides running work; search force-expands matching groups.
- public/prefs.ts: server-first load with a localStorage offline mirror, migrates
the legacy per-device proj-favs.
Backend (server-side prefs so favourites + collapse-state follow the user across
devices — localStorage was per-device, which broke the multi-device premise):
- src/http/prefs-store.ts: ~/.web-terminal-prefs.json store (mirrors push-store),
every field bounded + sanitized.
- GET /prefs (read-only) + PUT /prefs (Origin-guarded); UiPrefs type +
PREFS_STORE_PATH config.
Tests: +12 prefs-store, +21 groupProjects/displayLabel. Full suite 1227 passed.
Verified live: grouping renders, collapse persists to the server and survives a
full page reload (cross-device sync proven end-to-end).
The detail page now reads <repo>/CLAUDE.md and shows it in a scrollable
monospace box, with one smart button: "↻ Update" when it exists / "✨ Generate"
when it doesn't. The button opens an interactive Claude session in the repo
running /init (claude "/init") so you review what it writes — reuses the project
launcher, no new backend orchestration, stays a byte-shuttle.
- ProjectDetail += hasClaudeMd / claudeMd; buildProjectDetail reads CLAUDE.md
(truncated 64KB, best-effort, in the existing Promise.all)
- renderProjectDetail: CLAUDE.md section (content box or empty state) + button
- style.css: .proj-section-row / .proj-claudemd-btn / .proj-claudemd
Tests +5 (458): backend read present/absent; frontend content+Update,
empty+Generate, button opens `claude "/init"`. Verified: tsc clean, full suite
green, build OK, curl returns the content, in-browser shows the box + button.
Clicking a project's name opens an in-app detail view (back button returns to
the grid) with:
- header: name, path, branch, dirty
- Branch / Worktrees: each `git worktree list` entry (branch or detached@head,
main/current/locked/prunable tags, path) — a single-worktree repo shows its
branch; a multi-worktree repo shows them all
- Active sessions (N): detailed rows (status, devices watching, started-ago)
with Open + Kill
- the Claude/Codex/VS Code launcher row
Backend: src/http/worktrees.ts (parseWorktrees pure + listWorktrees via
execFile, no shell, best-effort); buildProjectDetail(cfg, path, liveSessions)
in projects.ts (validates absolute existing dir, reuses branch/dirty/session
helpers, uncached — detail is on-demand and wants fresh state); read-only
GET /projects/detail?path= (no Origin guard, like /projects; 400 missing /
404 invalid). Types: WorktreeInfo, ProjectDetail.
Frontend: detail view + navigation in projects.ts (grid<->detail), project
name becomes a link, Kill via DELETE /live-sessions/:id.
Tests +21 (452 total): worktree parser, buildProjectDetail (validation,
non-git, session merge, real git-init worktree), renderProjectDetail.
worktrees.ts 100% / projects.ts 93.8% cov. Verified in-browser: opened a
Claude session, then the detail showed its branch+worktree and the live
session with Open/Kill.
Replace each card's "+ start claude here" with a row of three brand-logo
launcher buttons (logo + caption, touch-friendly, tooltips):
- Claude (coral burst) → new terminal tab running `claude`
- Codex (OpenAI mark) → new terminal tab running `codex`
- VS Code (blue ribbon) → opens the repo in VS Code ON THE HOST
VS Code: new guarded POST /open-in-editor runs `<EDITOR_CMD> <path>` (default
`code`) on the host via execFile (no shell); path validated as an existing
absolute dir; Origin guard (CSRF) like the DELETE routes. EDITOR_CMD config added.
Active-session highlight (user request): when a project has a running Claude
session (Claude-hook status != unknown — plain shells/codex stay unknown), the
Claude logo gets a coral ring + tint, with a count badge when more than one.
New: public/icons.ts (simple-icons SVGs, fill=currentColor), src/http/editor.ts.
onOpenProject gains an optional cmd; makeProjectCard exported for tests.
Tests +16 (431 total): editor validation/launch, EDITOR_CMD config, launcher
row (3 buttons, claude/codex commands), and the active-Claude highlight + badge.
Verified: coverage >=80x4, build OK, both typechecks clean; in-browser the 3
logos render with brand colours, and POST /open-in-editor actually launched
VS Code on the host (403 no-Origin / 400 bad-path / 204 valid-dir all correct).
Prior-session prep for the v0.6 Project Manager, checkpointed on the
v0.6-projects branch as a clean base for the parallel builders:
- P1 (src/types.ts): ProjectInfo + ProjectSessionRef contracts; Config
gains projectRoots/projectScanDepth/projectScanTtlMs/projectDirtyCheck.
- P3 (src/config.ts): parseBool + parseProjectRoots helpers; parse the
4 PROJECT_* env vars with defaults ([homeDir]/4/10000/true).
- UI: public/style.css indigo -> Amber (#e3a64a) accent theme.
- Docs: FEATURE_PROJECT_MANAGER.md (1:N session design), THEME.md,
docs/mockups/ (final-amber.png is the locked visual).
Typecheck (backend + web) green. No behavior change to runtime code yet;
P2/P4/P5/P6 implement the feature on top of this.
The flat session list didn't tell you what each session was doing. Rebuild
/manage.html as a full-page grid where each card renders a LIVE preview
thumbnail — a read-only xterm showing the session's actual current screen,
scaled down like a screenshot — so you can recognize each at a glance.
- RingBuffer.tail(maxBytes): newest scrollback on chunk boundaries (no ANSI/UTF-8
split); GET /live-sessions/:id/preview returns {cols,rows,data} (no attach, so
it doesn't inflate watcher counts or keep sessions alive)
- manage.ts: card grid; per card a scaled read-only Terminal renders CLEAR+tail,
auto-refresh every 4s, click thumbnail/Open to join, Kill / bulk kill
- bundles xterm into manage.js (+manage.css linked in manage.html)
- ring-buffer tail tests
Verified: 3 sessions rendered as recognizable thumbnails (top / git log / ls),
colored, with status/age; click opens. 228 tests green, tsc clean.
A shared PTY can only be one size; min-sizing clamped the shared session to the
SMALLEST viewer, so a wide desktop got letterboxed when an iPad was also attached
(iPad looked full-screen, desktop did not).
Switch to latest-writer-wins: the device that most recently fit/focused drives the
PTY size, so whichever device you're actively using is full-screen.
- session.ts: setClientDims resizes the PTY directly (drop min across clients);
attach/detach/blur never resize (the active device keeps its size)
- frontend: TerminalSession.refit() force-resends dims; window 'focus' /
visibilitychange re-assert the active tab's size, so switching devices reclaims
full-screen
- tests updated for latest-wins (incl. join/hidden/blur don't resize)
Verified (two ws clients): 200x50 → iPad 100x40 → desktop refit 200x50 → iPad blur
keeps 200x50. 225 tests green.
The mirror looked broken because a backgrounded tab on one device still pinned
the shared PTY to its (default 80x24) size — so the device actively viewing got
a cramped terminal.
Fix: only an ACTIVELY-VIEWING client votes on PTY size.
- attachWs no longer seeds a default size vote (a join may be a hidden mirror)
- new 'blur' client message + clearClientDims(): a tab going hidden withdraws its
size vote (still mirrors output); show() re-casts it
- PTY size = min cols/rows across clients that have actually reported dims
- session/protocol tests cover hidden-mirror-doesn't-clamp + blur
Session manager (separate page, for the 'too many sessions' problem):
- GET stays; add DELETE /live-sessions/:id and DELETE /live-sessions[?detached=1]
- manager.killById(); LiveSessionInfo gains cols/rows
- public/manage.html + manage.ts: list/open/kill sessions, kill-all / kill-detached,
auto-refresh; 🗂 toolbar button; bundled as a 2nd esbuild entry
Verified: two concurrent clients mirror output + shared input; manage page
lists/kills (3→2→0). 225 tests green, tsc clean.
Relax the one-WS-per-session invariant (#5) so multiple devices mirror the
same live terminal (everyone sees output, anyone can type — tmux-style).
- types.ts: Session.attachedWs → clients Set + clientDims Map; add cwd;
SessionManager.list() + LiveSessionInfo
- session.ts: broadcast() output/exit to all clients; attachWs JOINS (no kick,
takes dims); detachWs removes one client, stamps detachedAt only when the last
leaves; setClientDims resizes PTY to the MIN cols/rows across clients
- manager.ts: join instead of kick; broadcast status; onExit drops only when a
client was attached; list() for discovery
- server.ts: GET /live-sessions; resize → per-client setClientDims; close →
detach one client; permission gate uses clients.size
- tests: rewrote session/manager tests for multi-client (join/broadcast/min-dims/
last-detach) + list(); 222 pass, tsc clean
- src/http/history.ts: listSessions() reads ~/.claude/projects/*/*.jsonl
(mtime-sorted, top 50), parseSessionMeta() extracts cwd + first user prompt
(pure, 4 unit tests); GET /sessions
- terminal-session: opts.initialInput (typed ~700ms after the shell is ready)
- tabs.newTabForResume(cwd,id): new tab in the project dir running
'claude --resume <id>'
- public/history.ts: 🕘 panel listing sessions (project · time · preview) with
Resume buttons
- Verified: /sessions returns 44 real sessions; panel lists project/time/preview.
216 tests green.
- title-util.cwdFromOsc7: parse cwd from OSC 7 (file:// URL); 4 unit tests
- terminal-session: register OSC 7 handler → cwd getter; opts.cwd → sent in the
attach frame on a fresh session
- protocol/types: ClientMessage attach.cwd (absolute path); validated
- session/manager/server: thread cwd → createSession spawn cwd (defaults homeDir)
- tabs: '+' (newTab) opens in the active tab's reported cwd, falls back to home
- Verified: server attach.cwd spawns there; cd /private/tmp → + → new tab pwd is
/private/tmp. 212 tests green.
- src/session/tmux.ts: sync tmux CLI wrappers (available/has/kill), best-effort
- config: useTmux from USE_TMUX env (1/0/auto→detect tmux on PATH)
- session: when useTmux, spawn 'tmux new-session -A -s web_<id> <shell>' (the
node-pty proc is a tmux CLIENT); createSession takes an optional id for
re-attach; Session.tmuxName; kill() runs tmux kill-session
- manager: handleAttach re-attaches to a surviving 'web_<id>' tmux session after
a restart (Case 3.5); shutdown kills only the client pty for tmux sessions so
the shell keeps running
- tests: USE_TMUX:0 in shared integration cfg (no tmux leakage); unit CFGs
useTmux:false; integration 'H1' (real tmux): set var → restart server →
re-attach → var survived. 208 tests green.
- types/protocol: ClientMessage approve/reject; ServerMessage status.pending
- server: POST /hook/permission HELD until the client decides; approve/reject
ws messages resolve it with {hookSpecificOutput.decision.behavior allow|deny};
5-min timeout + release-on-disconnect fall back to Claude's own prompt
- manager.handleHookEvent threads pending flag
- setup-hooks: PermissionRequest uses the held curl (writes decision to stdout);
status events stay fire-and-forget
- FE: terminal-session approve()/reject() + pending state; tabs approval banner
(Approve/Reject) for the active tab's held request
- Tests: protocol approve/reject; integration ⑧ held→approve→allow. 207 green.
- Browser-verified: banner 'Claude wants to use Bash' → Approve → resolves allow.
- All ARCHITECTURE §3 interfaces as pure types; dependency-free (no ws/node/DOM)
so both backend and frontend can import it
- Refinements vs §3 (documented): EnvLike (not NodeJS.ProcessEnv), WebSocketLike +
WS_OPEN (not ws.WebSocket), added Config.wsPath (invariant 8)
- ARCHITECTURE §3.1/§3.4 reconciled to match (anti-drift)
- tsc --noEmit passes
T2 of docs/PLAN.md
