Ctrl+C does not stop zeam binary started via lean-quickstart

[chetanyb]

Steps to reproduce

1. Build the zeam binary from the repo so zig-out/bin/zeam(or your custom location) exists and client-cmds/zeam-cmd.sh uses node_setup="binary".
2. Run NETWORK_DIR=local-devnet ./spin-node.sh --node zeam_0 --generateGenesis.
3. Wait for the node to start, then press Ctrl+C.

Behaviour

• Expected: the trap in spin-node.sh (cleanup + kill) should stop the zeam process and return the prompt.
• Actual: “exit signal received” + cleanup runs, but the zeam binary keeps logging and takes the terminal again. Cleanup repeats and eventually you have to kill it manually (Activity Monitor or pkill -f "zig-out/bin/
  zeam node"). Docker mode exits cleanly; the issue only happens with the binary mode set in client-cmds/zeam-cmd.sh.
• Sample log:

Dec-05 10:52:19.000 [info] (zeam): [chain] 
+===============================================================+
  CHAIN STATUS: Current Slot: -1 | Head Slot: 0 | Behind: 0
+---------------------------------------------------------------+
  Connected Peers:    0
+---------------------------------------------------------------+
  Head Block Root:    0x078f06926af6d9a214d5b234bab4514e27eec42f5ab2ffa0a6b473767aaa02d6
  Parent Block Root:  0x0000000000000000000000000000000000000000000000000000000000000000
  State Root:         0x9151351a9c269fa076a855ae79aa5a5e5552898ec581dd7d2ea78696dd7d1631
  Timely:             YES
+---------------------------------------------------------------+
  Latest Justified:   Slot      0 | Root: 0x078f06926af6d9a214d5b234bab4514e27eec42f5ab2ffa0a6b473767aaa02d6
  Latest Finalized:   Slot      0 | Root: 0x078f06926af6d9a214d5b234bab4514e27eec42f5ab2ffa0a6b473767aaa02d6
+===============================================================+

^Cexit signal received


cleaning up
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
docker rm -f zeam_0
Error response from daemon: No such container: zeam_0
kill -9 95617


cleaning up
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
./spin-node.sh: line 181: 95617 Killed: 9               eval "$execCmd"

docker rm -f zeam_0
Error response from daemon: No such container: zeam_0
kill -9 95617
./spin-node.sh: line 198: kill: (95617) - No such process

                                                                                                                                                                                                                        ▴3▿3
◄ 48s ⋈┈○ Dec-05 10:52:23.001 [info] (zeam): [chain]                                                                                                                                        □ lean-quickstart △⎪●◦◌◦⎥ 16:22
+===============================================================+
  CHAIN STATUS: Current Slot: 0 | Head Slot: 0 | Behind: 0
+---------------------------------------------------------------+
  Connected Peers:    0
+---------------------------------------------------------------+
  Head Block Root:    0x078f06926af6d9a214d5b234bab4514e27eec42f5ab2ffa0a6b473767aaa02d6
  Parent Block Root:  0x0000000000000000000000000000000000000000000000000000000000000000
  State Root:         0x9151351a9c269fa076a855ae79aa5a5e5552898ec581dd7d2ea78696dd7d1631
  Timely:             YES
+---------------------------------------------------------------+
  Latest Justified:   Slot      0 | Root: 0x078f06926af6d9a214d5b234bab4514e27eec42f5ab2ffa0a6b473767aaa02d6
  Latest Finalized:   Slot      0 | Root: 0x078f06926af6d9a214d5b234bab4514e27eec42f5ab2ffa0a6b473767aaa02d6
+===============================================================+

Notes

• I have encountered this bug on macos Tahoe 26.1, will add if I encounter this bug on other systems.
• Tested with other clients in binary mode like ream and this issue does not happen over there.