Email was the one sidecar built inside out. The platform held ~1,800 lines — the per-account
SQLite store, all 14 HTTP routes, account CRUD, resync, IMAP validation — while the 314-line
sidecar was a scheduler that reached BACK into the platform to do anything
(`import { performResync } from '../../api/email/resync'`).
The sidecar now serves its own HTTP listener and announces `email:server`, and
/api/email/* on the platform is createSidecarProxy like every other one: 1,801 lines down
to 22, with no mail knowledge left in it — not a message, not a folder, not a credential.
The routes moved verbatim, Hono and all. http.ts only reconstructs what the platform's
middleware used to provide: `user` on the context, from the X-Officer-User header the proxy
injects (trusted because this server binds loopback), and an error handler that turns
custom-errors into status codes.
The /email/events SSE stream went with them, which removes a whole round trip: the IDLE
watcher used to send `email:new` over the registration socket so the platform could push to
its SSE clients. Those clients are here now, so it calls broadcastEmailNew in-process and
`email:new` is gone from the wire protocol.
DELIBERATELY NOT DONE YET, and left backwards on purpose rather than half-moved:
- The two sync handlers (email-sync 381 lines, gmail-sync 712) still run in the platform's
queue and now import the store from its new home — a platform → sidecar import, which is
the wrong direction and is temporary. Moving them is option (A) from the plan: the sidecar
schedules its own syncs, independent of the platform Jobs list.
- accounts.ts still imports queue/init to enqueue a sync and to report sync status, and
index.ts still carries the queue-over-WS shim that inversion needs.
- The three channel handlers still open the mail store directly rather than asking over HTTP.
Two things worth knowing while testing: a from-scratch sync holds a proxied request open
well past the 60s idle default, hence timeoutSeconds on the proxy; and `gmail-sync` is
hardcoded in all three channel handlers even though the only account is provider=gmail with
auth_type=password, which routes to IMAP — so "sync emails" from a chat channel is
almost certainly already broken, and folds into the next stage.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
createSidecarProxy arrived with the wallet but nothing else moved onto it, so five sidecars
still carried their own copy of the same two files: a sidecar-server.ts that remembered a
port announced as `<name>:server`, and a router.ts that forwarded the subpath. Byte for
byte identical once the app name was normalised away — which is exactly what the factory's
own header said it existed to end.
headscale, transmission, invoiceshelf, slskd and music are now wallet-shaped: create the
proxy, export the router and the URL getter. 386 lines deleted against 163 added, and the
five feature directories go from ~70 lines each to ~18.
Two deviations were real and moved INTO the factory rather than being dropped, because both
are HTTP concerns rather than app knowledge:
- Range and If-None-Match are now forwarded for every sidecar. music needed both (seeking,
and ETag revalidation returning a cheap 304 instead of a cover image) and slskd needed
Range. Forwarding them everywhere costs nothing and removes the reason to hand-roll.
- timeoutSeconds, used only by music at 1800. A from-scratch reindex holds the proxied
connection open for minutes with no bytes flowing, which the 60s idle timeout would drop.
It applies to the whole prefix — the proxy must not know which of a sidecar's routes are
slow.
The five side-effect imports in hono.ts are gone with them: the port listener now registers
when createSidecarProxy runs inside the router this file already imports. Vault keeps its
hand-rolled pair and its side-effect import — it is off-limits by standing instruction, and
is the one sidecar this commit deliberately does not touch.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Terminals were a set of commands the platform drove. Officer sent pty:init / pty:input /
pty:resize / pty:close / pty:list over the registration socket, subscribed to ONE global
output stream, filtered every frame down to a session and rewrapped it — double
JSON-encoded — on the way out. That is terminal knowledge living in the process whose job
is authentication, and it made officer part of the data path for every keystroke.
The sidecar now serves its own loopback HTTP + WebSocket listener and announces the port
as `pty:server`, like every other HTTP sidecar. Officer authenticates the upgrade and
relays frames without reading them.
Split into three files, because "the sidecar" was one:
- sessions.mjs — the shell store. Spawn, attach, detach, resize, kill, scrollback, the
OSC-title scrape. Clients are a Set per session, so two panels can watch one shell.
- server.mjs — the listener. /ws speaks the browser's existing contract unchanged
({input,resize} in, {output,replay,exit,panel-refresh} out), plus /_officer/sessions,
DELETE /_officer/sessions/:id and POST /_officer/panel-refresh.
- index.mjs — the registration socket, and nothing else. It carries a port now.
On the platform side /api/terminal/* becomes createSidecarProxy, deleting the hand-rolled
router from two days ago, and websocket.ts drops from a translating bridge to a byte relay
modelled on the vault one. The whole PtyCommand/PtyEvent/PtyInitConfig/PtySessionInfo
vocabulary is gone from protocol.ts, connect.ts and sidecar-registry.ts.
broadcastPanelRefresh is now a POST to the sidecar: officer no longer holds terminal
sockets to loop over. Fire-and-forget — a missed refresh is a stale panel, not a failure.
The frontend did not move. The sidecar speaks what the browser already spoke.
The integration test was rewritten against the new shape, and tests something stronger than
before: officer is stopped mid-session and the shell keeps streaming, because officer is
not in the path at all. It also covers re-attach replay, the session list and kill.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Closing a terminal panel abandoned its shell. TerminalWrapper deleted the panel -> session
mapping on *unmount*, so any layout or route change generated a fresh uuid on the way back
and left the old shell running: alive, unreachable, and never killed, because nothing has
ever sent pty:close. The mapping now outlives the mount, so reopening a panel re-attaches
to the shell you left — which is also what finally makes the sidecar's replay buffer worth
having. It is persisted dashboard state, so this survives a reload too.
That trades an invisible leak for a visible one: a panel deleted for good still leaves its
shell behind. So `pty:list` now enumerates live sessions, and GET /api/terminal/sessions +
DELETE /api/terminal/sessions/:id expose them. pty:close finally has a sender.
Each session carries createdAt, lastActivityAt, pid, and the title the shell sets for
itself via OSC 0/2 — usually the running command, which is what turns "some uuid" into
"the one running claude" when you are deciding what to kill.
Killing on unmount is still not an option: it needs the panel system to distinguish a real
close from an incidental remount, which it cannot currently do.
Also raises the sidecar replay buffer from 50KB to 512KB — 50KB was about one long agent
turn, so reconnecting mid-task showed you the tail and nothing before it — and cuts the
buffer on a line boundary rather than a byte offset. A blind slice can land inside an
escape sequence, and the replay then opens with the tail of a colour or cursor-move code,
which xterm renders as garbage or applies as a real instruction.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
deletes the last of the projects/apps cluster: the published-app store
(/api/apps + /api/app-serve), the project dev-server and its websocket
proxy (/api/dev-server + /api/dev-server-proxy), the shared html-rewrite
they were the only consumers of, and their frontend — the Preview panel,
the UserApp panel/header, useUserApps and the /settings/apps screen.
also drops getUserProjectsDir and getUserAppsDir, the ProjectType and
ProjectDefinition types, and the 'dev-server' websocket provider from
server.tsx. nothing on disk is touched.
1440 deletions, 31 insertions. tsgo clean.
eight sidecars hand-rolled the same port capture — byte-identical once
the app name is normalised — and six repeated the same auth-and-forward
router. createSidecarProxy collapses both into one call and covers the
variants the others need: a ws:// url for music and vault, an onRegister
hook for opencode.
the wallet adopts it first: two files become one, 54 lines of router
become 16, and hono no longer needs a side-effect import to capture the
port. the no-body-parsing, no-body-logging rule moves into the factory
with its rationale, since that restraint is what keeps unlock
passphrases and macaroons out of the platform process.
costs 31 net lines today and pays back from the second adopter on.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
the owner's work, committed as one unit rather than split: the registration
files (App.tsx, Dock, AppRegistry, hono.ts, the schema and db barrels,
ecosystem.config.cjs) all reference modules under src/servers/{api,sidecar}/wallet
and src/workspaces/officerdev/src/apps/Wallet, so committing the shared plumbing
on its own would leave a commit that does not build.
officer-wallet is a new pm2 peer holding seed material sealed under an owner
passphrase on top of VAULT_STORE_KEY, with an unlock ttl after which the root key
is wiped from memory. five backends: on-chain via esplora, and lnd, clnrest,
lndhub and nwc for lightning. bolt11 encode/decode is implemented in-tree.
no secrets in the diff — the key-shaped literals under sidecar/wallet are the
bolt11 spec vectors and the bip39 "abandon … about" vector. .env.example gains
placeholders only. bun test src/servers/sidecar/wallet: 38 pass, 0 fail.
not reviewed line by line; assembled and verified to build, not audited.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
owns the invoiceshelf contract: instance url, sanctum token and the company
header that scopes every request. the platform side is the usual thin
auth+forward proxy at /api/invoiceshelf and holds no credentials.
built against the live 2.4.2 instance rather than the 3.0.0-alpha.1 checkout
in _references — the route allow-list came from artisan route:list on the
running container. they differ: 2.4.2 has estimates/{id}/convert-to-invoice
but no invoices/{id}/convert-to-estimate.
three upstream quirks absorbed here:
- accept: application/json is mandatory, or an unauthenticated request 302s
to an html login instead of returning 401
- origin/referer must never be sent, or statefulapi() switches to session+csrf
and every request 419s. the proxy forwards neither.
- a wrong company header does not error, it silently returns another company's
data. the pinned company is explicit and logged.
document pdfs are repaired: 2.4.2 prefixes them with a literal serialised http
response (201 bytes) inside a body already typed application/pdf. we slice to
the %PDF- magic. the report routes don't have the bug.
resources are an allow-list. backups, disks, modules, update/*, installation/*,
mail config, settings writes and ownership transfer stay unreachable, and the
per-resource action list keeps `send` — which really emails the customer —
from being reachable by accident.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
officer-transmission is a new pm2 peer that owns the transmission rpc
connection and exposes a curated /_officer/* contract instead of proxying
raw rpc. it absorbs the three quirks callers otherwise have to know about:
the 409 x-transmission-session-id handshake, failures returned as
{"result": "..."} inside http 200, and basic auth where an empty username
must send no header at all.
/transmission is the ui, on the workspace/panel framework: a filter nav and
three sections (torrents, stats, settings). the torrent list is virtualised
with 30 available columns, multi-select, and a right-click menu; the detail
pane covers general, files as a real tree, peers and trackers. filters and
the open torrent live in the url, so a filtered view is a link.
phase 1 goal was parity with _references/transmission-web. follow-up work is
recorded in TODO.md.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
officer-headscale owns the whole Headscale contract: the registered servers and
their admin api keys, the >=0.29 version floor, and every multi-call composition
the ui needs. the platform side is auth+forward only and holds no headscale
credentials, so the existing /api/vpn/enroll route and its HEADSCALE_* env vars
are untouched and unrelated.
officer manages many servers rather than one. the owner registers each with a url
and a key generated on that server and switches between them; exactly one is
active, enforced by a partial unique index rather than by convention. keys are
encrypted at rest and never leave the sidecar — the list projection cannot return
one. registration validates before it saves: an unauthenticated GET /version to
prove something headscale-shaped is there and meets the floor, then an
authenticated call to prove the key works. an edit that moves either half
re-validates.
there is deliberately no transparent /api/v1/* passthrough. headscale serialises
every uint64 as a json string and its rest shape moved repeatedly below 0.29;
proxying raw would push all of that into the browser, which is the mistake the
soulseek panels made with 37 raw upstream calls.
the /headscale workspace is nav + view over the panel system. only the servers
section is implemented — nodes, users and pre-auth keys say so plainly rather
than rendering an empty table that reads as a failed fetch.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The in-app Tailscale needs one thing from the platform: a way to turn an authenticated
Officer session into a Headscale pre-auth key, so the phone registers itself instead of
someone pasting a key by hand. The VPN's control and data planes talk directly to
Headscale — never through /api — so this is not a proxy and should not become one.
Headscale itself runs on a separate host, managed manually. The platform consumes
HEADSCALE_URL (also returned as controlUrl) and HEADSCALE_API_KEY, both from the host
env. Neither is set yet, which is why an unconfigured instance answers 503 rather than
crashing — Headscale is being stood up in parallel.
The response is `{ controlUrl, authKey }` exactly, because enrollVpn() in
@officer/core/officer-net.ts reads those two fields; changing the shape means changing the
app.
Two things the spec's sketch does not do:
- The `user` field changed meaning across Headscale versions — a name on <=v0.22, a
numeric id on v0.23+ — and we cannot see which one is running from here. So it resolves
the id via /api/v1/user and tries that first, falling back to the name. Whichever the
live server accepts wins, and neither version needs a config flag.
- Upstream calls carry a 10s timeout, and upstream error bodies are logged but never
returned to the client: that is an admin API and its errors are descriptive.
The standalone app's origin follows the platform's own convention rather than the spec's
literal: every other app origin is an env var with a scope rule, so this one is
OFFICER_TAIL_ORIGIN (set in .env on this host), restricted to /api/auth + /api/vpn the way
OffVault is restricted to /api/auth + /api/vault. The OffTail tile embedded in the main
Officer app needs nothing — it reuses OFFICER_APP_ORIGIN.
Not implemented: the optional GET/DELETE /api/vpn/devices. They are not needed for a first
connection and are better written against a running Headscale.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Being retired, so it is deleted rather than fixed — it had both defects dashboards just
had (edit dropping ?selected=, an abandoned edit following you onto the next item) and
repairing them was work for something on its way out.
Gone: the two screens and the panel apps, the three routes, the dock item and its entry
in the default dock paths (client and the three server-side copies), the page-title rule,
the app-registry entries, the barrel exports, the `projects` table with its queries and
row types, and the proj-meta/proj-layout/proj-terminals/proj-host-terminals branches in
the dashboards endpoint. The chat context and terminal state-key special cases for
`proj-layout-` went with them.
Deliberately kept: `getUserProjectsDir` in data-path.ts — the apps and dev-server routers
resolve user apps under the same on-disk Projects/ directory and are unrelated to this
feature. Nothing on disk is touched.
Needs `bun db:push` to drop the table; the schema change is the only thing standing
between the code and the database. One row was in it.
tsgo clean, 56/56 tests, no references left in src. Not yet exercised at runtime — the
restart is what will prove it.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Opening a chat from a non-default folder moved the cwd picker into a subdirectory of
that folder — three of the four sessions under ~/dockers/officer.dev jumped to
`platform`, the fourth to `Lyrics` — and everything after it was scoped to the wrong
directory. The list itself stayed correct, which is what made it confusing.
Two functions in claude-sessions.ts derived the cwd in opposite directions. parseSummary
(:103) keeps the first `cwd` it sees; parseClaudeTranscript (:182) overwrote it on every
entry, so it ended up with whatever a tool last cd'd into. A session's home is where it
was launched — that is how Claude Code files the transcript on disk and how the list
groups it — so the first one is right and the two now agree. It also stops a late entry
clobbering `fallbackCwd` when the caller already knew which group it was loading from.
Pre-existing, but only reachable since rows became real links (3682269): clicking one is
a route change now, so the deep-link resolver in ChatHistory/index.tsx:69 — previously
hit only on refresh or a pasted URL — runs on every click, and it feeds detail.cwd
straight into setActiveCwd.
Verified against the four real transcripts in that group: all four now resolve to
~/dockers/officer.dev. C4/C5/C6 pass in the browser (click, refresh, and fresh-tab
deep-link all land on the right folder).
Also adds docs/nav-test-checklist.md — the 56-check matrix for the navigation refactor
and the sidecar work, recovered from the transcript of the session that wrote it and
lost it (eff24773). The X block no longer needs its own branch now that sidecars is
merged. C is done, L/J/D/P/X/R are not.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
/email/messages and /email/stats built `labels LIKE '%${folder}%'` by string
interpolation, and `folder` comes straight off the query string. Five statements across
the two handlers were exposed. The fragment is bound now, and it carries its parameters
with it because each handler builds several statements from the same fragment and has
to spread them in order.
Checked against an in-memory table: inbox/INBOX/SENT/all return exactly what they
returned before, and `x' OR 1=1 --` now matches nothing instead of being SQL.
Also: page and limit reached the bindings as NaN for any non-numeric value, so a
mistyped query param was a 500. They fall back to their defaults now.
And deleted src/servers/sidecar/email-cron.ts — 92 lines imported by nothing. The live
cron is sidecar/email/email-cron.ts; this was an older copy that still reached into
queue-runner and google-auth directly, so leaving it there invites someone to fix the
wrong file.
This is the first commit on the email branch; the placement problems (the whole mail
store, both syncs, and the resync coalescing that cannot work across processes) are
untouched and much larger — see SIDECAR_WORK_LOG.md.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
cliamp playback was implemented entirely in officer: it located the cliamp binary,
validated the requested path against the owner's home, faked a PTY with `script`,
injected PULSE_SINK and an ALSA config shipped inside the API tree, spawned parec to
capture the sink, and set up the pulseaudio daemon and the virtual_out null sink at
every boot — about 356 lines of audio-pipeline knowledge in a process that is meant to
be a proxy, and none of it owned by the sidecar whose whole job is music.
all of it now lives in sidecar/music: cliamp-ws.ts serves both sockets (/cliamp/ws for
the player, /cliamp/audio/ws for the PCM capture) on the loopback server it already
runs, pulse-audio.ts does the daemon + sink setup at sidecar startup instead of at
officer's, and the asoundrc moved next to the code that passes it. officer keeps the
part that is actually its job — authenticating the browser — and relays frames both
ways without reading them (api/cliamp/relay.ts, same dumb-pipe shape as the vault
notifications relay). the browser's frame contract is unchanged, so the frontend is not
touched.
two things fixed on the way: the traversal check now requires a separator after the
home path, so a sibling directory whose name merely starts with it can no longer pass;
and the music proxy no longer special-cases /reindex and /reindex/stream by name to
extend the idle timeout — it extends the whole prefix, because a proxy should not know
which of the sidecar's routes are slow.
the music-specific `files` query param is out of the shared WS envelope too: upgradeWs
now carries the raw query string, which any relayed provider can use.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The second copy of the same problem. The opencode sidecar reported raw
ChatEvents and officer translated them, buffered the assistant text and wrote
every durable message to chat_session_events — so an officer restart mid-turn
lost whatever the model had produced since the last write, and `connect.ts`
dropped the events that arrived while it was down without a word.
Both harnesses speak ChatEvents, so the sidecar reuses the agent's session log
verbatim: translate, commit, then deliver the finished message with its cursor
id as `opencode:message`. Officer folds it into the in-memory transcript and
relays it, exactly as it now does for claude — `createEventHandler` (166 lines,
a duplicate of turn-stream.ts) and `emitToSession` are gone, and nothing in
officer writes to chat_session_events any more.
`opencode:event` stops being a wire event; it is the runner's internal report to
the sidecar it runs in, typed as such so it cannot leak back onto the socket.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The password lives in the owner's ~/.vnc, and the sidecar is the process that
writes it — together with the rfbauth file x11vnc actually authenticates
against. Officer read the plaintext half directly and answered with it before
ever asking the sidecar, which is both a secret the proxy has no business
opening and a way to hand out a password that no longer matches: if `passwd`
went missing while `password` survived, officer kept serving the old plaintext
and the desktop refused every login. `vnc:ensure-password` reconciles the pair,
so ask it every time and delete the reader.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
it was the only sidecar living outside src/servers/sidecar/ — it sat in
api/terminal/ next to the bridge that talks to it, which is the one place a reader
looking for "the sidecars" would not check. now src/servers/sidecar/pty/index.mjs,
matching every peer, with a note on the pm2 entry about why this one is node and
.mjs (node-pty is a native addon) rather than bun and typescript like the rest.
the templates/ directory went to api/users/, next to provision.ts:seedShellConfigs,
which is now its only consumer — the sidecar's duplicate seeder went with the
sandbox branch in the previous commit. api/terminal/ is left holding exactly one
thing: the websocket bridge.
no behaviour change. the pm2 entry's script path changed, so `pm2 restart
officer-pty` is not enough — pm2 remembers the old path until the entry is deleted
and started again. commands are in SIDECAR_WORK_LOG.md.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
officer built the whole PtyInitConfig: it read the owner's SHELL (defaulting to
/bin/zsh), added `-i`, read their HOME, expanded `~` against it, and hardcoded
`host: true`. none of that is a proxy's business — the sidecar is the process that
calls pty.spawn, so it is the one that should know what to spawn and where.
the config now carries only what the bridge actually knows: sessionId, the folder
the panel was opened on, and the client's cols/rows. shell, args, home and cwd
resolution moved into the sidecar. home comes from HOME_DIR ?? HOME, mirroring
data-path.ts:getOwnerHomeDir — terminal was the one host-executing surface reading
process.env.HOME directly, which is identical here and divergent anywhere HOME_DIR
is set to something else.
deleted the bwrap sandbox branch rather than moving it. it was selected by
`config.host`, which officer hardcoded to true, so it never ran — and it expected
`shell` to contain a fully-built bwrap command that nothing on either side ever
built. it could not have worked. a terminal here is the owner's own shell on the
owner's own machine by design (platform/CLAUDE.md), so there is no jail to preserve.
its ensureUserFiles half duplicated api/users/provision.ts:seedShellConfigs, which
is the live seeder of those same templates and stays.
also deleted the 'cwd' handler that turned a message into `cd <path>\r` typed at
the shell. no frontend has ever sent that message — the browser composes its own cd
— so it was unreachable, and synthesizing keystrokes is not something a relay
should do.
the integration test pins SHELL and HOME_DIR now that the sidecar reads them, and
asserts the shell starts in the resolved `~` rather than officer having resolved it.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
the pty sidecar registered `term.onData` with the socket that happened to be live
when the session was created. officer is a pm2 peer that restarts constantly, and
every restart hands this process a brand new socket, so every pre-existing session
went on writing to a closed one — where sendJson's readyState check dropped it
silently. the shell survived and still accepted input, because input arrives on the
new socket, but nothing ever came back. you typed and the terminal sat there. the
only way out was to close the panel, which orphaned the shell.
sendJson now reads the module-level socket at send time instead of taking one as an
argument, so there is no socket to capture and go stale. that is the whole fix.
the scrollback replay on re-attach becomes its own event, pty:replay -> 'replay'
on the browser socket. it used to arrive as ordinary output, which was fine for a
page load (fresh xterm) but not for a restart: the browser keeps its terminal, so
replaying blind printed a second copy of everything still on screen. marked as
history, Terminal.tsx resets and rebuilds from the sidecar's 50KB buffer instead.
it also stays out of the `output` branch so it cannot re-trigger the command /
initial-input logic that scrapes output for a sentinel.
added an integration test, because this is a reconnect bug and nothing short of an
actual reconnect proves it: it stands up a fake registration socket, runs the real
sidecar against it, echoes into a real shell, kills the socket, rebinds the same
port the way pm2 does, and asserts output still flows. verified it fails against
the old sendJson (times out after 15s waiting for the post-restart echo) and
passes in ~400ms with the fix. it never touches the running officer — the sidecar
dials API_URL, overridden per spawn.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
officer's registration socket silently drops sends when it isn't OPEN
(sidecar/connect.ts:send — no queue, no error, no return value). the agent pushed
raw parser events over that socket and officer translated and persisted them, so
everything a turn produced while officer was restarting went nowhere: the turn kept
running, the output was gone, and a reconnecting client replayed a log that simply
had no rows for those seconds. stage 1 kept the agent alive across a restart; this
is what makes its output survive one too.
move the translation and the write into the sidecar:
- turn-stream.ts is the stateful ChatEvent -> browser-message translator lifted out
of websocket.ts (delta buffering, flush before tool:start and result). pure and
synchronous, so it is unit tested — 12 tests, 100% lines.
- session-log.ts commits each message to chat_session_events and only then hands it
to officer, with its cursor id attached. per-session promise chain: translation is
synchronous and therefore in arrival order, and only the commit is queued, so
cursor ids are assigned in the order events actually happened. a delta that
overtook the assistant:text in front of it would make the client commit its stream
buffer at the wrong point, so deltas go through the same queue even though they are
never written.
- claude:event on the wire becomes claude:message: a finished browser-facing message
plus its seq. officer relays it verbatim and folds it into the in-memory session
for sync:messages. it no longer builds or persists chat messages for this harness.
gap detection, which is what the durable log is for. chat_session_events.id is a
global bigserial, so two consecutive events of one session are not consecutive ids
and a client cannot tell a contiguous replay from one with a hole in it. each durable
message now carries prevSeq — the cursor of the previous message in the same session —
which is inside the persisted payload, so it survives replay. useChat compares it
against the cursor it holds before advancing, and surfaces a visible marker on a
mismatch: a conversation that silently skips a tool call or half an answer reads as
the assistant having done something inexplicable. only checked once a cursor exists,
because opening a session from history legitimately starts mid-chain (events are swept
after 7 days, the transcript is not).
a failed write delivers live with no seq, so the client sees the message but does not
advance past something it cannot replay, and the next successful write chains from the
cursor the client still holds.
pipeline steps pass durable: false. their sessionKey is a throwaway uuid no browser
will ever replay and the job's own event log is its record, so writing those rows only
grows the table.
opencode still goes through officer's createEventHandler, now labelled as such. that
is the sidecars-opencode branch.
this fixes R4 from CLAUDE_SIDECAR_ISOLATION.md. R3 and R5 already worked.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
the process that runs claude (sidecar/claude/user-instance.ts) had no pm2 entry
and was spawned on demand by the main server, with stdout/stderr inherited. that
made every agent session a grandchild of officer, so pm2's tree-kill took the
session down on every `pm2 restart officer` — the single thing that makes it
impossible to work on the platform while an agent is running.
give it its own entry (officer-agent) and delete the spawn machinery:
ensureClaudeSidecar, spawnAndWaitForRegistration, the 50ms registration poll and
the per-email claudeProcs/claudeSpawnWaiters maps, ~77 lines. officer now spawns
no sidecar at all.
for that to work the sidecar had to stop needing officer to start:
- it resolves the owner from the database (getOwnerUser) instead of reading
CLAUDE_USER_EMAIL out of the env officer built. single-user is a hard
invariant, so there is nothing to fan out over. CLAUDE_USER_EMAIL still wins
when set, for manual runs, and a fresh install waits for bootstrap rather
than exiting into a restart loop.
- it reads the anthropic proxy secret from the proxy sidecar's own state file
rather than being handed it in env. lazily, because ensureProxySecret
persists on a 30s debounce and pm2 starts both processes together.
it registers as 'agent' with capability 'claude', so the registry finds it the
way it finds every other sidecar. that removes the email argument from
killClaude, interruptClaude and clearClaudeSession, which only ever existed to
locate a per-email sidecar by name.
what officer keeps is a short wait-for-capability, because pm2 brings peers up
together and the first request after a boot can beat the sidecar's registration.
also align the two officer port fallbacks in the sidecar (5000 for the socket,
9010 for the rest base) — same instance, so they cannot disagree.
this fixes R1 and R2 from CLAUDE_SIDECAR_ISOLATION.md. events produced while
officer is down are still lost; that is stage 2.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
extremely long transcripts made bottom-anchoring the virtualized list unreliable
(thousands of unmeasured variable-height items = a huge estimate the scroll never
lands on). now GET /chat/sessions/:id takes limit+before and returns a windowed
slice plus total+offset. the chat opens on the last 20 messages, anchors to the
bottom instantly, and scrolling near the top pages in the next older window,
prepending it and pinning the previously-top message so the view stays put.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
deep-linking or refreshing /chat/<id> only has the id, so add loadClaudeSessionById
to scan every project group for the transcript and return its real cwd. the session
list page resolves on load from that, setting the cwd picker and resuming. session
rows are now <Link to=/chat/<id>> so clicking updates the url and flows through the
same resolve path.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
slskd has no favorites or buddy-list concept — 0.26.0's UsersController exposes
only endpoint/browse/directory/info/status — so Officer owns that data itself:
a soulseek_favorites table served by the slskd sidecar under a /_officer/*
namespace, which can never collide with slskd's /api/v0/*. The main server
gains exactly one line, injecting X-Officer-User on the proxy hop, so it stays
a thin auth proxy and grows no Soulseek logic. The route is handled before the
upstream check, so favorites keep working with slskd down.
Usernames in search results and downloads become a dropdown (browse shares,
toggle favorite). Browsing publishes to a nonce-stamped, consumed-once channel
so the Users section looks the peer up without re-running the expensive browse
on every remount, and favorites get their own section at the top of that panel,
which doubles as its landing content. CardHeader had to split its toggle row to
host the dropdown, since a trigger can't live inside the collapse button.
The schema file is deliberately self-contained so it can move wholesale into
the sidecar directory when sidecars start owning their own schema. Its DDL was
applied by hand, matching drizzle's constraint naming, rather than running a
whole-schema push.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add officer-slskd, a singleton sidecar that reverse-proxies to a
self-hosted slskd (Soulseek) instance and reports its loopback port to
the API on connect. All slskd knowledge (URL + API key) lives in the
sidecar; the platform is a thin auth+forward proxy for /api/slskd/* and
holds no slskd credentials. Mirrors the officer-vault pattern.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The crypto-shape question is answered (classic + v2 fields both present); drop
the diagnostic that logged the response field names.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Vaultwarden (2025.12.0/f21a3ada) returns both classic (Key/PrivateKey/Kdf*) and
v2 crypto (UserDecryptionOptions.MasterPasswordUnlock, AccountKeys) synthesized
from the classic stored fields. session/login now returns the whole native
connect/token response minus the transport tokens (under `connectToken`),
alongside the spec's protectedUserKey/privateKey/kdf aliases, so the SDK gets
whatever unlock path it uses. Temp diagnostic logs the response field NAMES
(values redacted) to empirically confirm on a real login.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- token-store: shared "give me a valid Vaultwarden access token" (proactive
refresh) used by both the HTTP proxy and the WS; router refactored onto it.
- notifications WS: validates the platform session in `open` (deferred, owner
only), injects the stored Vaultwarden token into the upstream, and buffers
client frames during the async setup so the SignalR handshake isn't dropped.
The device connects with its platform JWT (?access_token=), never a vault one.
- lifecycle: logout drops the vault token set (keeps the protector); distress
(/auth/revoke) and panic wipe both token set and protector, forcing a
one-time master-password re-setup.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Implements the platform half of VAULT_AUTH_SPEC.md. /api/vault is now gated on
an owner platform session (userMiddleware, no bodyParser → streaming preserved)
and origin-scoped as before; the device holds no Vaultwarden token.
- POST /session/login {email, authHash, kdf, device*} → broker calls Vaultwarden
/identity/connect/token via the sidecar, stores the encrypted token set tied to
the owner, and returns {protectedUserKey, privateKey, kdf} (ciphertext to us).
- GET/PUT /unlock-key → store/release the Officer-app protector key (owner only).
- Catch-all proxy swaps the incoming platform JWT for the stored Vaultwarden
access token, proactively refreshes near expiry, and retries once on a 401 for
replayable requests. Bodies are never parsed.
client_id column added to vault_tokens (needed to refresh). Broker error text is
read across Vaultwarden's message/errorModel/error fields.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Transparent pass-through fronting a self-hosted Vaultwarden so the OffVault
(Bitwarden-SDK) app reaches it through the platform's per-app origin gate. True
out-of-process sidecar (officer-vault): it owns all Vaultwarden knowledge (URL,
paths, notifications WebSocket) on a random loopback port and registers via the
sidecar connector; the platform is a thin origin-gated forwarder that knows only
the sidecar's port. Never decrypts/parses/rewrites/logs bodies.
- sidecar/vault: HTTP + notifications-WS proxy to VAULTWARDEN_URL, /_health
- api/vault: sidecar-port discovery + thin forwarder + WS pipe + origin gate
- origin: OFFICER_VAULT_ORIGIN allow-listed, scoped to /api/vault
- mounted top-level (not protected) so the Bitwarden bearer token isn't 401'd
- protocol: vault:server event; ecosystem: officer-vault app
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
move video/audio downloads off the dedicated ReClip download lane onto the
generic script-job path:
- delete execute-download.ts, reclip-client.ts and the POST /jobs/download
endpoint; drop the 'download' mode from the pipeline_jobs enum (legacy rows
tolerated)
- execute-script.ts: strip the @@officer:progress@@ sentinel from the log,
emit progress events, and isolate viewer/log writes (safeEmit/safeLog) so a
broadcast or log throw can't wedge the stdout pump
- pipeline-job-manager.ts: persist latest progress; guard sendToViewer sends
- ScriptJobDetail: render the two progress bars; DownloadJobDetail kept for
legacy history rows
- TaskRunnerModal: ScriptRunner descends into the triggered directory
- VideoDownloadPanel: rewire startJob to the script-job path
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A failed download has no media file, so its description was lost — no way to tell
which video was missed. Now a failure writes a `<title>.txt` (or `<videoId>.txt`
when untitled) holding the URL + description, so every missed item leaves a
recoverable reference. Always written (even with an empty description — the URL is
the reference); skipped only on a deliberate Stop, not a genuine error.
Verified: successful item → "<media base>.txt" (description); failed item →
"<title>.txt" (url + description); spaces preserved in both.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The metadata phase already fetches ReClip's description (ReClip now forwards it in
/api/info). Carry it into the download phase and write it to a text file with the
same base name as the media — "Song Name.mp3" → "Song Name.txt".
reclipDownloadOne gains an onFilename callback that fires the moment the final
filename is known (before the file transfers), so the executor writes the sidecar
in parallel with the download stream, and the exact name guarantees they pair up.
Empty descriptions write nothing; the write is best-effort (never fails a download).
Verified: correct base name + .txt, exact content, and no sidecar for an empty
description.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Verified against live ReClip: POST /api/download with title:"" → a hash filename
(b5d04adc86.mp3); with title:"Me at the zoo" → "Me at the zoo.mp3". So ReClip
names the file from the title WE send (falling back to a hash) — it does not
self-name. The title is mandatory, which means a metadata pass is required.
Back to two phases:
1. metadata — fetch each item's /api/info (title + validity), keep survivors, skip
errors.
2. download — download each survivor passing its title, so files land with real
names; skip download errors.
Keeps the exact-urls[] input (Mix playlists can't drift) and the one-request-per-
item download. Progress is two counters again (Titles + Download); UI shows two
bars. ~2 requests/item is inherent to needing the title (per the user's call:
correctness over speed).
Verified two-phase filtering + title passthrough + skip-on-error with a mock.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A YouTube Mix/radio playlist (list=RD…) returns a different set of items on every
/api/playlist call (observed 779 / 1485 / 529 for the same URL). The job used to
re-expand the playlist server-side, so it would download a different list than the
count shown on the decision screen.
The panel now passes the already-expanded `urls[]` into the job, and the executor
uses them verbatim (falling back to expanding `url` only when no list is given).
The job downloads exactly what you decided on. Endpoint takes `urls[]` (stored as
inputs.urls) or `url`.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The job's two phases were a misread — the "count" phase is the client-side
playlist expansion (for the inline-vs-job decision, already done in the panel).
The job itself is just one download request per item.
Dropped the in-job metadata pass entirely:
- reclip-client: reclipDownloadOne no longer prefetches /api/info for a title —
ReClip names the file from the video title itself, so it's a single request
per item.
- execute-download: one phase — expand the playlist, then /api/download each url,
skip failures. Progress is a single { done, failed, total, current } counter
(no meta/dl split); ~2× faster and downloads start right after expansion.
- UI (DownloadJobDetail + panel JobView): one "Downloaded" bar instead of two.
Verified: every item is attempted directly (no /api/info gate), skip-on-error
counts correct.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Front half of the download-job feature.
Panel (VideoDownloadPanel): after Fetch expands a playlist and the count is known,
a decision screen — "Found N items" → pick Audio/Video + subfolder → "Download all
as a job" (POST /jobs/download), or "fetch inline to pick individually" (the
existing card grid). A single video still goes straight to the inline card. The
job phase shows live two-phase progress (polled from the job) + a "View in Jobs"
link; it notes the job runs server-side so closing the panel is fine, and it
refreshes the browser as each file lands.
/jobs (DownloadJobDetail + JobsPage dispatch): a `download` job renders a compact
two-phase readout — Metadata and Download bars (processed/total, found/skipped and
saved/failed) + the current item — polled from the job's progress, with a Stop.
Executor tweak: phase-1 meta.done now counts kept (not processed) so both phases
read the same `(done+failed)/total`.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Turns the downloader into a server-side job on the existing jobs spine (Postgres
persistence, live WS viewers + replay, abort, /jobs UI) — but with its own
executor and its own lane, since it's deterministic scripting, not an agent, and
a multi-hour playlist mustn't block agentic jobs.
- reclip-client.ts (new, shared): reclipInfo / reclipPlaylist / reclipDownloadOne
(single download → streams the file to a dir, abort-aware). Extracted so both
the file-browser endpoints and the job executor use one client.
- execute-download.ts (new): the two-phase executor —
phase 1 metadata (expand playlist, fetch each info, keep survivors, skip
errors), phase 2 download (each survivor in the chosen format; skip download
errors). Emits a compact `download:progress` snapshot (counters, not per-item
events — playlists are thousands of items). Throws on abort / fatal.
- job manager: `download` mode dispatch → executeDownload; persists
download:progress; adds execution LANES (download vs default) so the two run
independently and each serializes on its own; promoteNext fills both lanes.
- POST /api/tasks/jobs/download { url, format, dir, root?, label? } — enqueues a
download job (own lane, no capability task needed; traversal-guarded target).
- schema: `download` added to the mode enum (drizzle text-enum — no DB migration);
getPendingJobs() query for lane filling.
Verified the executor with a mocked ReClip client: two-phase filtering, skip-on-
error counts, and abort-throws all correct.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Two problems behind "deleted/changed the folder image but it still shows" (on
both web and app):
1. Client caching. Cover/meta/etc. were served with an ETag(=v) but NO
Cache-Control, so browsers served them straight from the heuristic cache at
the same URL — a changed cover kept showing the old image. And the platform
proxy never forwarded If-None-Match, so the ETag revalidation couldn't work
anyway. Now the sidecar sends `Cache-Control: no-cache` on every version-
stamped artifact (cover/meta/poster/lyrics/image) + the manifest, and the
proxy forwards If-None-Match → the client revalidates every time and gets a
cheap 304 when unchanged, a fresh 200 when v changed.
2. Removed covers lingered. On rebuild the indexer only (over)wrote cover.jpg
when a source cover existed — a deleted or now-undecodable source left the
old cover.jpg in the cache (still served, still cover:true). Now it clears
cover.jpg first and regenerates only if there's a valid source.
Verified live against a booted sidecar: cover carries no-cache + ETag, a
matching If-None-Match → 304, and deleting the source folder.jpg drops the
cached cover (404, meta.cover cleared, manifest cover:false).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Reworks the download-video dialog into a prefetch-then-download flow, mirroring
ReClip's own web UI. Still a pure proxy to ReClip (its yt-dlp); no downloader
logic moves to the platform.
Server (thin ReClip proxies alongside /download-video):
- POST /file-browser/video-info { url } → ReClip /api/info → { title, thumbnail, duration, uploader }
- POST /file-browser/video-playlist { url } → ReClip /api/playlist → { urls }
Both return { error } inline (200) so the client can render failures per-card.
UI (VideoDownloadDialog, now self-contained; useFileBrowserApp exposes `files`
and drops the old single-shot state/handler):
- Paste a URL → Fetch. A playlist URL (list=) expands via /video-playlist, then
each entry's /video-info is prefetched sequentially (ReClip does yt-dlp per
video), rendering a card (thumbnail, title, uploader, duration) that fills in
progressively.
- Per-entry Download, plus Download All when there's more than one; per-card
status (downloading → saving → saved / retry-on-error) via the existing
background job + poll.
- Playlists get an optional "subfolder you name" field (ReClip's /api/playlist
carries no playlist title); blank = current folder.
- Quality is always best (matches the mobile Share flow — no picker); the
audio-only toggle applies to the whole batch.
Verified ReClip's contract live: /api/info returns the metadata fields, and
/api/playlist returns { urls } (17 entries in ~1.2s).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Favorites / now-playing / playlists were being served by the platform router
straight from Postgres, which violated the intended split (officer = auth +
proxy; officer-music = the whole /api/music/* contract). Move them into the
sidecar so it owns ALL music endpoints — library AND user state.
- sidecar (index.ts): serves /favorites, /now-playing, /playlists[/:id[/items]]
backed by Postgres (the same officerdb queries other sidecars already use).
The authenticated user id arrives in X-Officer-User; the sidecar is loopback-
only, so it trusts the header (401 if absent). HTTP-contract comment updated.
- platform (router.ts): reduced to a pure auth+proxy catch-all — it now injects
X-Officer-User from the authenticated ctx user and forwards the request body
(favorites/now-playing/playlist writes carry JSON) in addition to Range/query.
No schema change — the tables are unchanged, only WHERE they're served moves.
Verified live: booted the real sidecar against the live DB and exercised the
endpoints with the X-Officer-User header — 401-without-header, favorites round-
trip, and full playlist CRUD with ownership scoping all pass.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Mirrors the existing per-user music state (favorites / now-playing): Postgres
tables + query layer + REST endpoints on the music router, all scoped to the
caller's user id and served directly by the platform (not proxied to the
user-stateless sidecar). Item `key`s are opaque track homePaths, same contract
as favorites — the server never interprets them.
- schema: music_playlists (name unique per user) + music_playlist_items
(0-based position, dupes allowed, cascade delete).
- queries: get/create/rename/delete playlists; add (append) / set (replace,
covers reorder+remove) items; every mutation ownership-checked; item ops in a
transaction that also bumps the playlist updatedAt.
- router (/api/music, before the catch-all proxy): GET/POST /playlists,
GET/PATCH/DELETE /playlists/:id, POST/PUT /playlists/:id/items. 409 on
name collision, 404 on a playlist that isn't the caller's.
- migration 0002 (also backfills music_favorites/now_playing into the snapshot,
which were originally applied via a direct db:push). Applied to the DB.
Verified end-to-end against the live DB: create, dupes, ordering, append,
replace/reorder, ownership scoping, rename, counts, delete — all pass.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Display-only panel at the top of the Get Lyrics run-task form (single file only):
- Backend: GET /file-browser/audio-meta?path= — ffprobe format tags + duration,
plus a second probe for embedded lyrics (USLT/SYLT/lyrics* keys, case-insensitive).
Returns { title, artist, duration, hasLyrics }; tolerant of missing tags/probe
failures.
- Client: files.audioMeta(path) + AudioMeta type in useFilesAPI.
- TaskRunnerModal: prefetch audioMeta for get-lyrics single-file runs (bypasses
the hasTrackPickers early-return, error-tolerant), and render AudioMetaPanel
above TaskInputForm — title/artist + a muted length + Lyrics: Yes/No chip,
filename fallback. Directories + other tasks unaffected (no panel, no probing).
Verified ffprobe logic on a real embedded-lyrics file. tsgo clean; formatted.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Reading the manifest (which every app refresh hits) used to call
ensureIndexFresh(), kicking off a debounced rebuild — and after the CACHE_VERSION
bump that meant a plain refresh could launch a full library rebuild. Make reads
side-effect-free: /manifest now just returns the last completed index. Builds are
explicit only (POST /reindex or the SSE stream); pick up disk changes by
reindexing.
Removes the now-unused ensureIndexFresh + lastBuildFinishedAt, and drops
/manifest from the 30-min per-request timeout extension (both hops) since it no
longer blocks on a build.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
pastilhasandClaude Opus 5