A message starting with `?` in a mapped room runs `claude -p` one-shot in that
repo on the Mac and posts the full answer back into the room — Matrix as a
request/response interface, not just a trigger. Non-`?` messages keep launching
interactive sessions as before.
New scripts/ask-claude.sh is a login-shell wrapper (so ~/.zprofile puts claude on
PATH) that exports CLAUDE_CODE_OAUTH_TOKEN from the Mac's .env and runs
`claude -p "$prompt" < /dev/null`, printing the answer to stdout. The bot adds a
`?`-dispatch with run_ask/ask: SSH stdout captured, 300s timeout, fail-loud, output
chunked under Matrix's event cap (no truncation).
Headless claude -p needs the long-lived token because a non-GUI SSH session can't
reach the login Keychain (reports "Not logged in") — the deliberate Approach A that
the interactive GUI-Terminal path (D11) avoided. Token is kept Mac-side only; the
Spark never runs claude. Sovereignty unchanged: claude -p uses the subscription, no
frontier API touches message payloads.
Proven live on the Spark; fresh-eyes reviewed before commit.
Add Dockerfile, docker-compose.yml, docker-entrypoint.sh, and .dockerignore
so the bot runs detached and survives reboots, replacing the foreground venv run.
The image is generic (no secrets/deployment specifics baked in): host networking
reaches both Synapse and the Mac; .env, config.toml, and the SSH key are mounted
read-only. The entrypoint is the container's environment seam (D4 analog of
launch-claude.sh) — it generates ~/.ssh/config for the mac-bridge alias from
config.toml [mac] (new hostname/user fields) so the bot's `ssh mac-bridge` stays
unchanged. SSH key mounted not baked; first connect uses accept-new host trust.
Proven live on the Spark: container connects to Synapse and real messages launched
drivable sessions on the phone across 2 rooms via the full chain.
Keysat