stablyai/orca · warning · Error

${response.error.message}

Error message

${response.error.message}

What it means

Inside the NewWorktreeModal, a useEffect fires `ssh.getState` for the selected repo connection to read the current SSH connection state when the modal becomes visible. If the RPC returns ok=false, it throws the host's error.message; the chained .catch() maps that into an sshState with status 'error' and the thrown message as the error field.

Source

Thrown at mobile/src/components/NewWorktreeModal.tsx:413

    })()
    return () => {
      stale = true
    }
  }, [visible, client, hostId])

  useEffect(() => {
    if (!visible || !client || !selectedRepoConnectionId) {
      return
    }
    let stale = false
    void client
      .sendRequest('ssh.getState', { targetId: selectedRepoConnectionId })
      .then((response) => {
        if (stale) {
          return
        }
        if (!response.ok) {
          throw new Error(response.error.message)
        }
        const state = (response as RpcSuccess).result as { state?: SshConnectionState | null }
        setSshState(
          state.state ?? {
            targetId: selectedRepoConnectionId,
            status: 'disconnected',
            error: null,
            reconnectAttempt: 0
          }
        )
      })
      .catch((err) => {
        if (!stale) {
          setSshState({
            targetId: selectedRepoConnectionId,
            status: 'error',
            error: err instanceof Error ? err.message : 'Failed to read SSH connection state.',
            reconnectAttempt: 0

View on GitHub (pinned to 1136503c6a)

Solutions

1. Verify the SSH target still exists on the host (re-open repo settings).
2. Upgrade the host to support ssh.getState.
3. Let the modal's error state guide the user; the catch already sets status 'error'.
4. Retry by closing/reopening the modal (the effect re-fires on visible toggle).

Example fix

// before
if (!response.ok) {
  throw new Error(response.error.message)
}

// after — treat missing-method as disconnected rather than error
if (!response.ok) {
  if ((response as RpcFailure).error.code === 'METHOD_NOT_FOUND') {
    setSshState({ targetId: selectedRepoConnectionId, status: 'disconnected', error: null, reconnectAttempt: 0 })
    return
  }
  throw new Error((response as RpcFailure).error.message)
}

// Skip the probe if the host likely lacks ssh.* methods
const caps = await client.sendRequest('status.get').catch(() => null)
const supportsSsh = (caps as any)?.result?.capabilities?.includes('ssh')
if (!supportsSsh) { setSshState({ targetId: selectedRepoConnectionId, status: 'disconnected', error: null, reconnectAttempt: 0 }); return }

function isSshStateSuccess(r: RpcSuccess | RpcFailure): r is RpcSuccess & { result: { state?: SshConnectionState | null } } {
  return r.ok
}

.then((response) => {
  if (stale) return
  if (!response.ok) throw new Error((response as RpcFailure).error.message)
  // ...
}).catch((err) => {
  if (!stale) setSshState({ targetId: selectedRepoConnectionId, status: 'error', error: err instanceof Error ? err.message : 'Failed to read SSH connection state.', reconnectAttempt: 0 })
})

When it happens

Trigger: ssh.getState fails — the SSH target was deleted, the host's SSH subsystem errored, the connection id is unknown to the host, or the host predates the ssh.getState method.

Common situations: Selecting an SSH-backed repo whose target was removed server-side, a host restart that lost SSH state, an older host without ssh.* methods, or a transient relay error.

Related errors

• Unable to create untitled markdown note
• Unable to load workspace metadata.
• Unable to reach host
• Unable to load agent sessions
• ${response.error.message}

AI-assisted analysis of stablyai/orca@1136503c6a (2026-08-12). Data as JSON: /api/errors/b213ecfacf73e2d0. Report an issue: GitHub.