Checks
Operating system
Linux 5.14.0-611.47.1.el9_7.x86_64 - Rocky Linux 9.7 (Blue Onyx)
Expected behaviour
Installing via curl -fsSL https://cli.kiro.dev/install | bash on a remote SSH server should either result in a working kiro-cli chat session, or produce a clear error explaining what is missing and what configurations are actually supported. kiro-cli doctor should accurately diagnose the real cause of each failure rather than printing generic remediation that doesn't match the actual error condition.
Actual behaviour
kiro-cli chat exits immediately with code 1, producing no output, no error message, and no indication of where logs might be found:
$ kiro-cli chat; echo "exit=$?"
exit=1
kiro-cli doctor reports the wrong cause for an unreadable sshd_config. On Rocky 9 the file is mode 600, so the check fails on EACCES, but instead of surfacing that the output prints a directive-checklist as if the file content were wrong:
✘ sshd config: Could not read sshd_config
The /etc/ssh/sshd_config file needs to have the following line:
AcceptEnv Q_SET_PARENT
AllowStreamLocalForwarding yes
After sudo chmod 644 /etc/ssh/sshd_config the check passes immediately with no directive changes, confirming the failure was purely a permission issue. Doctor then moves on to a second failure:
✘ Kiro CLI terminal integrations: kiro-cli-term is not running in this terminal, please try restarting your terminal
Restarting the terminal, reconnecting the SSH session, and reinstalling shell integrations all have no effect. On a headless remote there is no desktop parent process for the wrapper to attach to, and no amount of restarting changes that. The advice to restart the terminal is not actionable in this setup.
Additionally, the diagnostic output shows install-method = "unknown" even though the install was done via the official curl -fsSL https://cli.kiro.dev/install | bash script, suggesting that install path is not being detected.
Finally, this issue template asks me to confirm I ran q doctor and q restart, and to search the amazon-q-developer-cli issue tracker for duplicates. The product is called Kiro CLI, the binary is kiro-cli, and there is a separate cli.kiro.dev domain — but Q_SET_PARENT in sshd_config, desktop-release.q.us-east-1.amazonaws.com as the download host, and the issue template itself all still reference the old name. It's unclear to users whether old q-related documentation and issues apply.
Steps to reproduce
On a Rocky Linux 9 host with no desktop environment, accessed only via SSH, run curl -fsSL https://cli.kiro.dev/install | bash and follow the prompts. Open a fresh SSH session and run kiro-cli chat — it exits immediately with code 1 and no output. Run kiro-cli doctor — it fails on the sshd_config check with the misleading directive error described above, and after fixing permissions it fails again on the terminal-integration check with no working remediation.
Environment
<This will be visible to anyone. Do not include personal or sensitive information>
[q-details]
version = "2.2.2"
hash = "7b8bf7d8e854fbf3b4071b047526b3dd851b2563"
date = "2026-05-05T23:02:24.043902Z (2d ago)"
variant = "minimal"
Note: TERM=tmux-256color in the diagnostic is the terminal type configured on this host — this session was not inside a tmux session. Q_SET_PARENT_CHECK=1 is present, suggesting the shell integration ran, but the actual parent process socket is not available in this headless environment.
[system-info]
chip = "QEMU Virtual CPU version 2.5+"
total-cores = 7
memory = "13.39 GB"
[system-info.os.linux]
kernel_version = "5.14.0-611.47.1.el9_7.x86_64"
id = "rocky"
name = "Rocky Linux"
pretty_name = "Rocky Linux 9.7 (Blue Onyx)"
version_id = "9.7"
version = "9.7 (Blue Onyx)"
[environment]
cwd = "/home/USER/projects/uselist"
cli-path = "/home/USER/projects/uselist"
os = "Linux"
shell-path = "/usr/bin/zsh"
shell-version = "5.8"
install-method = "unknown"
in-ssh = true
[env-vars]
PATH = "/home/USER/.opencode/bin:/home/USER/.local/bin:/home/USER/.bun/bin:/home/USER/.local/bin:/home/USER/.pyenv/shims:/home/USER/.pyenv/bin:/home/USER/.local/share/codex-dual/bin:/home/USER/.pyenv/versions/3.13.12/bin:/home/USER/.pyenv/libexec:/home/USER/.pyenv/plugins/python-build/bin:/home/USER/.pyenv/plugins/pyenv-virtualenv/bin:/home/USER/.pyenv/plugins/pyenv-update/bin:/home/USER/.pyenv/plugins/pyenv-doctor/bin:/home/USER/.opencode/bin:/home/USER/.local/bin:/home/USER/.bun/bin:/home/USER/.local/bin:/home/USER/.pyenv/bin:/home/USER/.opencode/bin:/home/USER/.local/bin:/home/USER/.bun/bin:/home/USER/.nvm/versions/node/v24.14.0/bin:/home/USER/.local/bin:/home/USER/.pyenv/bin:/usr/local/bin:/usr/bin:/usr/local/sbin:/usr/sbin:/home/USER/bin:/home/USER/bin"
Q_SET_PARENT_CHECK = "1"
SHELL = "/usr/bin/zsh"
TERM = "tmux-256color"
XDG_SESSION_TYPE = "tty"
Checks
q doctorin the affected terminal sessionq restartand replicated the issue againOperating system
Linux 5.14.0-611.47.1.el9_7.x86_64 - Rocky Linux 9.7 (Blue Onyx)
Expected behaviour
Installing via
curl -fsSL https://cli.kiro.dev/install | bashon a remote SSH server should either result in a working kiro-cli chat session, or produce a clear error explaining what is missing and what configurations are actually supported. kiro-cli doctor should accurately diagnose the real cause of each failure rather than printing generic remediation that doesn't match the actual error condition.Actual behaviour
kiro-cli chat exits immediately with code 1, producing no output, no error message, and no indication of where logs might be found:
kiro-cli doctor reports the wrong cause for an unreadable sshd_config. On Rocky 9 the file is mode 600, so the check fails on EACCES, but instead of surfacing that the output prints a directive-checklist as if the file content were wrong:
✘ sshd config: Could not read sshd_config
The /etc/ssh/sshd_config file needs to have the following line:
After sudo chmod 644 /etc/ssh/sshd_config the check passes immediately with no directive changes, confirming the failure was purely a permission issue. Doctor then moves on to a second failure:
✘ Kiro CLI terminal integrations: kiro-cli-term is not running in this terminal, please try restarting your terminal
Restarting the terminal, reconnecting the SSH session, and reinstalling shell integrations all have no effect. On a headless remote there is no desktop parent process for the wrapper to attach to, and no amount of restarting changes that. The advice to restart the terminal is not actionable in this setup.
Additionally, the diagnostic output shows install-method = "unknown" even though the install was done via the official curl -fsSL https://cli.kiro.dev/install | bash script, suggesting that install path is not being detected.
Finally, this issue template asks me to confirm I ran q doctor and q restart, and to search the amazon-q-developer-cli issue tracker for duplicates. The product is called Kiro CLI, the binary is kiro-cli, and there is a separate cli.kiro.dev domain — but Q_SET_PARENT in sshd_config, desktop-release.q.us-east-1.amazonaws.com as the download host, and the issue template itself all still reference the old name. It's unclear to users whether old q-related documentation and issues apply.
Steps to reproduce
On a Rocky Linux 9 host with no desktop environment, accessed only via SSH, run
curl -fsSL https://cli.kiro.dev/install | bashand follow the prompts. Open a fresh SSH session and runkiro-cli chat— it exits immediately with code 1 and no output. Runkiro-cli doctor— it fails on the sshd_config check with the misleading directive error described above, and after fixing permissions it fails again on the terminal-integration check with no working remediation.Environment