- startRunAction serializes per (workspace, action); a double restart no longer orphans a run.
- After approval, re-read orca.yaml and run exactly the approved action (never a stale cache).
- Host restarts snapshot running actions (arm/relaunch/disarm) so tab-list rebuilds can't drop
them; relaunch skips the interrupt and keeps the old tab if approval is declined.
- Port registry re-reads the shared file before allocating and refreshes use daily (LRU eviction).
- OSC 133;D without an exit code reads as exited; action names are single-line.
- Cloud wake hook moved to orca-cloud/cloud-run-action-relaunch.ts.
- orca.yaml scripts.run.<id> {command, name, cwd, env, default, runMode, availableIn},
scripts.runMode (concurrent|nonconcurrent), scripts.autoRunAfterSetup; snake_case and the
bare-string form are accepted for Conductor parity.
- Every workspace gets a stable, collision-free block of 10 ports per host: ORCA_PORT,
ORCA_PORT_COUNT and a CONDUCTOR_PORT alias, in every terminal and the setup/archive env.
- Run button + Cmd/Ctrl+Alt+Enter: runs open a 'Run: <name>' terminal; Stop interrupts, then
closes the terminal after 5 s; Restart; status from OSC 133; detected ports link to Ports.
- Shared run actions go through orca.yaml trust ('run' kind); imported ones are local.
- Conductor importer maps run/run_mode/auto_run_after_setup (settings.toml, settings.local.toml,
conductor.json runScriptMode) instead of skipping them.
- Cloud layer: relaunch running actions after wake and rollover.
W10 wake-visibility P1s. stderr '- …' lines (NDJSON phase events with
--progress json; --json defaults to none; stdout keeps only the result):
- wake: 'Waking <ws>...', one 'Still waking (Ns)…' line at 10 s,
'<name> is awake (took Ns).', or, when the API answers still waking
after its wait, '<name> is still waking. Check with: …' (exit 0, status
waking kept in the result)
- sleep: 'Putting <ws> to sleep...' / '<name> is asleep (took Ns).'
- worktree create --on cloud: 'Still starting (Ns)…' at 10 s and
'Workspace <id> is running (took Ns).'
Lifecycle handler moved to cloud-lifecycle.ts; shared helper cloud-progress.ts.
A user's own orcad started from a server pane must not share or overwrite the server's
marker, so the variable joins ORCAD_SERVER_LAUNCH_ENV_KEYS. The launcher may name a directory
that does not exist yet; create it (0700) before writing the marker.
The dispatcher now takes either a method list or a loader. The RPC server's default is a loader
that imports the method registry on the first call; orcad preloads it right after its ready line,
so a reconnecting client normally finds it loaded. Concurrent first calls share one load; a failed
import answers those calls with an RPC error and the next call retries.
The legacy relay CLI imported the registry statically, which kept it on orcad's startup path; it
now loads it with its first relayed command (its dispatcher moves to its own file for that).
Measured on a 4-vCPU VM from the current golden, n=20 interleaved: readyMs p50 679 -> 621,
p95 716 -> 665.
The desktop's runtime record kept saying 'running' after a sleep from another
client until lost-contact tripped (minutes), so an open or predictive wake
returned at once without resuming, and the row read awake.
- Wake path: a cloud record still 'running' asks the control plane; when the
sandbox is suspended the record is marked asleep and the resume runs.
- Status reads (window refresh, predictive preflight, the 1 min beat) mark a
running record asleep when the control plane says sleeping, so the row
reads Asleep. Only when the record is unchanged since it was read.
- Quiet (predictive) wakes report whether a resume actually ran; one that
found the VM up is 'skipped': no unopened hold, no backoff, no short idle
window in main.
- A skipped wake (preflight no, or nothing to resume) gives the 30 s
hover/focus gap back.
#151 added the /proc foreground proof, but the OSC 133 prompt hint only reaches 'prompt'
after a clean C -> D;0 -> A cycle, so the first command typed into a new terminal (every
wake's first command) still waited the 500 ms paste settle. A runtime mirror that starts
at the shell's spawn treats its first prompt as clean; any C, failed D or seed (restored or
attached stream) keeps the old rule. Linux only, where the /proc proof gates it.
terminal.send text+Enter to a daemon-backed shell always waited the 500 ms agent paste
settle: the daemon's shell-foreground answer is lifecycle ownership, which is only proven
after a recovery or full-screen exit, never for a fresh or plain-command shell. On Linux
the idle-prompt claim now also accepts the shell's own /proc stat (process group is the
terminal foreground group, i.e. ps '+'), scoped to the target PTY; other hosts unchanged.
The acceptance wake split uses one persistent paired connection per wake (the desktop's
transport) by default; transport=cli keeps the per-call socket the CLI dials.
- Forward tunnel early data (workspace-ports.forward-tunnel.early-data.v1). It is negotiated on
each attach: the client asks for it and the host echoes it in `ready`. The client sends up to
64 KiB, its receive credit and a half-close right behind Open. The host holds those bytes
until the destination connects, writes them before Opened, and drops them if the connect
fails. Old hosts and old clients keep the strict ordering, and the browser tunnel is unchanged.
New-connection HTTP GET through a forward on a real VM: p50 127 -> 64 ms, p95 138 -> 70 ms.
- SSH and port-forward idle-sleep activity now use ForwardUseMeter instead of the
2 KiB/min byte threshold. An idle VS Code remote window sends 2.3-3.0 KB/min of
keepalives, so it kept the VM awake forever. The meter counts packet bursts and
payload-sized packets instead; the unit tests replay traces recorded on the real VM.
- Downloads: files.stat now goes out with the first pages instead of one round trip ahead
of them.