ops: remove bridge tmux guidance, keep tmux for agent sessions (#140)

Author: benvinegar

bin/baudbot

@@ -351,16 +351,6 @@ case "${1:-}" in
     shift
     require_root "restart"
     if has_systemd; then
-      # Ensure any pre-existing detached bridge tmux session is torn down so
-      # restart always boots a fresh bridge from currently deployed runtime files.
-      AGENT_USER="${BAUDBOT_AGENT_USER:-baudbot_agent}"
-      if command -v tmux >/dev/null 2>&1; then
-        if command -v sudo >/dev/null 2>&1; then
-          sudo -u "$AGENT_USER" tmux kill-session -t slack-bridge 2>/dev/null || true
-        elif command -v runuser >/dev/null 2>&1; then
-          runuser -u "$AGENT_USER" -- tmux kill-session -t slack-bridge 2>/dev/null || true
-        fi
-      fi
       exec systemctl restart baudbot "$@"
     else
       echo "systemd not available."

bin/baudbot.test.sh

@@ -130,7 +130,7 @@ EOF
   )
 }
 
-test_restart_restarts_systemd_and_kills_bridge_tmux() {
+test_restart_restarts_systemd() {
   (
     set -euo pipefail
     local tmp fakebin log_file
@@ -168,12 +168,6 @@ if [ "${1:-}" = "-u" ]; then
 fi
 echo "sudo $*" >> "${BAUDBOT_TEST_LOG}"
 exec "$@"
-EOF
-
-    cat > "$fakebin/tmux" <<'EOF'
-#!/bin/bash
-echo "tmux $*" >> "${BAUDBOT_TEST_LOG}"
-exit 0
 EOF
 
     cat > "$fakebin/systemctl" <<'EOF'
@@ -182,11 +176,10 @@ echo "systemctl $*" >> "${BAUDBOT_TEST_LOG}"
 exit 0
 EOF
 
-    chmod +x "$fakebin/id" "$fakebin/sudo" "$fakebin/tmux" "$fakebin/systemctl"
+    chmod +x "$fakebin/id" "$fakebin/sudo" "$fakebin/systemctl"
 
     PATH="$fakebin:$PATH" BAUDBOT_TEST_LOG="$log_file" BAUDBOT_ROOT="$tmp" bash "$CLI" restart
 
-    grep -q '^tmux kill-session -t slack-bridge$' "$log_file"
     grep -q '^systemctl restart baudbot$' "$log_file"
   )
 }
@@ -198,7 +191,7 @@ run_test "version reads package.json" test_version_uses_package_json
 run_test "status dispatches via runtime module" test_status_dispatches_via_runtime_module
 run_test "attach requires root" test_attach_requires_root
 run_test "broker register requires root" test_broker_register_requires_root
-run_test "restart kills bridge tmux then restarts systemd" test_restart_restarts_systemd_and_kills_bridge_tmux
+run_test "restart restarts systemd" test_restart_restarts_systemd
 
 echo ""
 echo "=== $PASSED/$TOTAL passed, $FAILED failed ==="

bin/lib/baudbot-runtime.sh

@@ -398,7 +398,7 @@ cmd_attach() {
         echo "  sudo baudbot attach                  # defaults to control-agent"
         echo "  sudo baudbot attach --pi control-agent"
         echo "  sudo baudbot attach --pi <uuid>"
-        echo "  sudo baudbot attach --tmux slack-bridge"
+        echo "  sudo baudbot attach --tmux sentry-agent"
         exit 0
         ;;
       *)

pi/skills/control-agent/SKILL.md

@@ -339,14 +339,12 @@ The sentry-agent operates in **on-demand mode** — it does NOT poll. Sentry ale
 
 ### Starting the Slack Bridge
 
-The `startup-cleanup.sh` script handles bridge (re)start automatically — it detects broker vs Socket Mode, reads the control-agent UUID, and launches the bridge in a `slack-bridge` tmux session.
+The `startup-cleanup.sh` script handles bridge (re)start automatically — it detects broker vs Socket Mode, reads the control-agent UUID, and starts the bridge as a normal background process.
 
-If you need to restart the bridge manually:
+If you need to restart the bridge manually, rerun startup cleanup and then inspect logs:
 ```bash
-MY_UUID=$(readlink ~/.pi/session-control/control-agent.alias | sed 's/.sock$//')
-tmux kill-session -t slack-bridge 2>/dev/null || true
-tmux new-session -d -s slack-bridge \
-  "unset PKG_EXECPATH; export PATH=\$HOME/.varlock/bin:\$HOME/opt/node-v22.14.0-linux-x64/bin:\$PATH && export PI_SESSION_ID=$MY_UUID && cd ~/runtime/slack-bridge && exec varlock run --path ~/.config/ -- node broker-bridge.mjs"
+bash ~/.pi/agent/skills/control-agent/startup-cleanup.sh UUID1 UUID2 UUID3
+tail -n 200 ~/.pi/agent/logs/slack-bridge.log
 ```
 
 Verify: `curl -s -o /dev/null -w '%{http_code}' -X POST http://127.0.0.1:7890/send -H 'Content-Type: application/json' -d '{}'` → should return `400`.
@@ -364,7 +362,7 @@ If you need to check manually, use `heartbeat trigger` to run all checks immedia
 When the heartbeat reports a failure, take the appropriate action:
 1. **Missing sentry-agent**: Respawn with tmux and re-send role assignment.
 2. **Orphaned dev-agents**: Kill tmux session and remove worktree.
-3. **Bridge down**: Restart the `slack-bridge` tmux session.
+3. **Bridge down**: Restart via `startup-cleanup.sh`, then check `~/.pi/agent/logs/slack-bridge.log`.
 4. **Stale worktrees**: `git worktree remove --force` + `rmdir` empty parents.
 5. **Stuck todos**: Escalate to user via Slack.
 

pi/skills/control-agent/startup-cleanup.sh

@@ -7,7 +7,7 @@
 # Pass the live session UUIDs (from list_sessions) as arguments.
 # Any .sock file whose UUID is NOT in the live set gets removed.
 # Stale .alias symlinks pointing to removed sockets also get cleaned.
-# Then restarts the slack-bridge tmux session with the current control-agent UUID.
+# Then restarts the slack-bridge process with the current control-agent UUID.
 
 set -euo pipefail
 
@@ -66,11 +66,39 @@ else
   exit 1
 fi
 
-# Kill existing slack-bridge tmux session if running
-if tmux has-session -t slack-bridge 2>/dev/null; then
-  echo "Killing existing slack-bridge session..."
-  tmux kill-session -t slack-bridge
+BRIDGE_PID_FILE="$HOME/.pi/agent/slack-bridge.pid"
+BRIDGE_LOG_DIR="$HOME/.pi/agent/logs"
+BRIDGE_LOG_FILE="$BRIDGE_LOG_DIR/slack-bridge.log"
+
+kill_bridge_supervisor() {
+  local bridge_pid="$1"
+  [ -n "$bridge_pid" ] || return 0
+  if ! kill -0 "$bridge_pid" 2>/dev/null; then
+    return 0
+  fi
+
+  # Best-effort: terminate direct children first so no stale bridge process keeps the port.
+  local bridge_child_pids
+  bridge_child_pids="$(pgrep -P "$bridge_pid" 2>/dev/null || true)"
+  if [ -n "$bridge_child_pids" ]; then
+    kill $bridge_child_pids 2>/dev/null || true
+    sleep 1
+    kill -9 $bridge_child_pids 2>/dev/null || true
+  fi
+
+  kill "$bridge_pid" 2>/dev/null || true
   sleep 1
+  kill -9 "$bridge_pid" 2>/dev/null || true
+}
+
+# Kill existing slack-bridge process if running
+if [ -f "$BRIDGE_PID_FILE" ]; then
+  BRIDGE_PID="$(cat "$BRIDGE_PID_FILE" 2>/dev/null || true)"
+  if [ -n "$BRIDGE_PID" ] && kill -0 "$BRIDGE_PID" 2>/dev/null; then
+    echo "Killing existing slack-bridge process (pid=$BRIDGE_PID)..."
+    kill_bridge_supervisor "$BRIDGE_PID"
+  fi
+  rm -f "$BRIDGE_PID_FILE"
 fi
 
 # Select bridge script: prefer broker pull mode when SLACK_BROKER_* vars are present,
@@ -100,9 +128,25 @@ if [ -z "$BRIDGE_SCRIPT" ]; then
 fi
 
 # Start fresh slack-bridge
+# Keep a supervisor loop (matching start.sh) so bridge restarts automatically on crash.
 echo "Starting slack-bridge ($BRIDGE_SCRIPT) with PI_SESSION_ID=$MY_UUID..."
-tmux new-session -d -s slack-bridge \
-  "unset PKG_EXECPATH; export PATH=\$HOME/.varlock/bin:\$HOME/opt/node-v22.14.0-linux-x64/bin:\$PATH && export PI_SESSION_ID=$MY_UUID && cd /opt/baudbot/current/slack-bridge && exec varlock run --path ~/.config/ -- node $BRIDGE_SCRIPT"
+mkdir -p "$BRIDGE_LOG_DIR"
+(
+  unset PKG_EXECPATH
+  export PATH="$HOME/.varlock/bin:$HOME/opt/node-v22.14.0-linux-x64/bin:$PATH"
+  export PI_SESSION_ID="$MY_UUID"
+  cd /opt/baudbot/current/slack-bridge
+  while true; do
+    varlock run --path ~/.config/ -- node "$BRIDGE_SCRIPT" >>"$BRIDGE_LOG_FILE" 2>&1
+    echo "[$(date -Is)] ⚠️  Bridge exited ($?), restarting in 5s..." >>"$BRIDGE_LOG_FILE"
+    sleep 5
+  done
+) &
+NEW_BRIDGE_PID=$!
+echo "$NEW_BRIDGE_PID" > "$BRIDGE_PID_FILE"
+chmod 600 "$BRIDGE_PID_FILE"
+echo "Bridge pid: $NEW_BRIDGE_PID"
+echo "Bridge logs: $BRIDGE_LOG_FILE"
 
 # Wait for bridge to come up
 sleep 3

start.sh

@@ -83,16 +83,37 @@ fi
 
 if [ -n "$BRIDGE_SCRIPT" ]; then
   RELEASE_BRIDGE="/opt/baudbot/current/slack-bridge"
-  tmux kill-session -t slack-bridge 2>/dev/null || true
-  echo "Starting Slack bridge ($BRIDGE_SCRIPT)..."
-  tmux new-session -d -s slack-bridge \
-    "export PATH=$HOME/.varlock/bin:$HOME/opt/node-v22.14.0-linux-x64/bin:\$PATH && \
-     cd $RELEASE_BRIDGE && \
-     while true; do \
-       varlock run --path ~/.config/ -- node $BRIDGE_SCRIPT; \
-       echo '⚠️  Bridge exited (\$?), restarting in 5s...'; \
-       sleep 5; \
-     done"
+  BRIDGE_LOG_DIR="$HOME/.pi/agent/logs"
+  BRIDGE_LOG_FILE="$BRIDGE_LOG_DIR/slack-bridge.log"
+  BRIDGE_PID_FILE="$HOME/.pi/agent/slack-bridge.pid"
+
+  mkdir -p "$BRIDGE_LOG_DIR"
+
+  # Stop any previous bridge process tracked by pid file.
+  if [ -f "$BRIDGE_PID_FILE" ]; then
+    old_pid="$(cat "$BRIDGE_PID_FILE" 2>/dev/null || true)"
+    if [ -n "$old_pid" ] && kill -0 "$old_pid" 2>/dev/null; then
+      kill "$old_pid" 2>/dev/null || true
+      sleep 1
+      kill -9 "$old_pid" 2>/dev/null || true
+    fi
+    rm -f "$BRIDGE_PID_FILE"
+  fi
+
+  echo "Starting Slack bridge ($BRIDGE_SCRIPT)... logs: $BRIDGE_LOG_FILE"
+  (
+    export PATH="$HOME/.varlock/bin:$HOME/opt/node-v22.14.0-linux-x64/bin:$PATH"
+    cd "$RELEASE_BRIDGE"
+    while true; do
+      varlock run --path ~/.config/ -- node "$BRIDGE_SCRIPT" >>"$BRIDGE_LOG_FILE" 2>&1
+      echo "[$(date -Is)] ⚠️  Bridge exited ($?), restarting in 5s..." >>"$BRIDGE_LOG_FILE"
+      sleep 5
+    done
+  ) &
+  # Intentionally track the supervisor subshell PID (not per-restart node child PID)
+  # so a single kill stops the entire bridge restart loop.
+  echo $! > "$BRIDGE_PID_FILE"
+  chmod 600 "$BRIDGE_PID_FILE"
 fi
 
 # Set session name (read by auto-name.ts extension)