How to Configure Windsurf SSH Permissions
You set up windsurf ssh permissions by limiting the local SSH keys your editor can use. You also restrict what target servers allow for each login key or certificate. Note that on June 2, 2026, Cognition renamed Windsurf to Devin Desktop. The app keeps the same IDE features and uses its agent to run terminal commands. When teams leave SSH keys wide open, the agent inherits full server reach. A bad prompt or mistake can then change or delete live files. How windsurf ssh permissions work in the
SSH Access Management + AI Agent Permissions: the complete guide
Architecture patterns, implementation strategies, and security best practices. Delivered to your inbox.
You set up windsurf ssh permissions by limiting the local SSH keys your editor can use. You also restrict what target servers allow for each login key or certificate. Note that on June 2, 2026, Cognition renamed Windsurf to Devin Desktop. The app keeps the same IDE features and uses its agent to run terminal commands. When teams leave SSH keys wide open, the agent inherits full server reach. A bad prompt or mistake can then change or delete live files.
How windsurf ssh permissions work in the terminal
In Windsurf, now called Devin Desktop, the agent runs terminal commands with user approval. It opens a dedicated shell, using zsh on macOS or bash on Linux. That shell loads your normal profile, aliases, and variables. Because it runs as your local user, it can reach your local SSH agent. Any remote host your shell can log into is open to the coding agent.
To control these sessions, teams rely on native OpenSSH controls on the target hosts. Instead of static keys, you can sign user keys with an OpenSSH Certification Authority. OpenSSH added certificate support for users and hosts in version 5.4. An OpenSSH certificate uses a simpler format than X.509 certificates. It holds a public key, identity text, principal names, and signed options.
When you sign a user key, you set a key identity and list allowed principals. Principals are the user or host names that the certificate lets in. You can also set a time window so the signed certificate expires on schedule. If you skip the time window during signing, the certificate stays valid forever. By giving the agent a dedicated key with strict options, you limit where it logs in.
What native SSH setup solves and what it misses
Native OpenSSH certificates and principal rules solve the initial login problem well. They let your team verify which machine or user opened a remote connection. You can issue short-lived certificates so stolen keys stop working after their window ends. You can also map the coding agent to a low-privilege Linux user on the server.
Still, host login rules only check who starts the session at the door. Once the SSH session opens, the remote shell runs whatever commands the Linux user can run. Host keys do not inspect individual shell commands while the agent works in flight. If the Linux role can restart services or read secret files, the agent can too.
SSH Access Management + AI Agent Permissions: architecture patterns and best practices
Continue reading? Get the full guide.
Free. No spam. Unsubscribe anytime.
Local approval prompts inside the IDE help, but engineers get tired of reading long scripts. A single approved command can still pull down private data or alter a live host. Also, standard SSH logs show that a user logged in, not every command output. When an AI agent runs dozens of commands across hosts, auditing those actions gets hard. Teams also struggle when terminal outputs contain emails, phone numbers, card numbers, or national IDs. Without inline checks, sensitive text flows right back into the agent context window.
Steps to tighten local and remote SSH rules
Start by separating your personal admin keys from the shell that your editor uses. Keep high-privilege keys out of the default key ring that loads in your shell profile. That way, the agent cannot start a root session on a live server by mistake.
Next, use a Certificate Authority to sign short-lived user certificates for specific tasks. Include only the exact principal names needed for the dev or staging host. Always set a clear validity window when signing so old certificates do not pile up. Finally, pair those host rules with command-level checks before letting agents touch live systems.
Add command and file guardrails to SSH access
Host firewalls and certificate principals control who opens an SSH session. Yet an allowed user or agent can still run a bad one-shot command or pull sensitive files over sftp. In practice, hoop.dev sits inline on the host as an open-source SSH server with guardrails that uses OpenSSH certificates. For every one-shot command or sftp task an agent or human runs in production, hoop.dev decides to pass, mask, block, or escalate it for review.
Using hoop.dev SSH controls, guardrails check one-shot commands, env variable names, and sftp or scp paths before they run. Port forwards default to deny unless the target is listed. Because interactive shells have no statement boundary and carry no guardrails, teams can turn them off so every command is checked and logged. Check out the MIT-licensed project on GitHub.
Frequently asked questions
Did Windsurf change its name to Devin Desktop?
Yes, Cognition announced that Windsurf became Devin Desktop on June 2, 2026. All old Windsurf docs now point to Devin Desktop, and the IDE features stay the same.
How does the Windsurf agent use local SSH keys?
The agent runs commands in a dedicated zsh or bash shell with user permission. Because that shell loads your user profile, it can use your active SSH keys and variables.
Can OpenSSH certificates limit which user accounts an agent uses?
Yes, when you sign an OpenSSH certificate, you list allowed principal names and validity times. The remote server checks those signed principals before it lets the SSH login proceed.
Related guides
In this post