How to Configure Devin SSH Approval
You set up devin ssh approval by requiring human review before an agent logs in. Your team can pair agent prompts with short-lived OpenSSH certificates for each task. Many teams start by giving an AI agent a static SSH key. That key often holds broad rights across many servers. If the agent makes a mistake, a broad key lets it break critical files. Strict approval rules stop an agent from running unchecked commands on your hosts. How devin ssh approval works with OpenSSH AI coding agents need
Approval Chains & Escalation + SSH Access Management: the complete guide
Architecture patterns, implementation strategies, and security best practices. Delivered to your inbox.
You set up devin ssh approval by requiring human review before an agent logs in. Your team can pair agent prompts with short-lived OpenSSH certificates for each task. Many teams start by giving an AI agent a static SSH key. That key often holds broad rights across many servers. If the agent makes a mistake, a broad key lets it break critical files. Strict approval rules stop an agent from running unchecked commands on your hosts.
How devin ssh approval works with OpenSSH
AI coding agents need shell access to check logs, test code, and fix bugs. In a safe setup, the agent does not hold a permanent private key. Instead, the agent asks for permission when a task needs server access. A human engineer reads the request and checks the target host name.
Once the engineer approves the request, your signing tool creates an OpenSSH user certificate. OpenSSH added support for user and host certificates in version 5.4. These certificates use a much simpler format than standard TLS certificates. A certificate authority key signs the public key of the agent. The signed certificate holds a key identity string and a list of allowed principals. Principals are the specific user or host names the agent may use.
You can also set a tight validity window on the certificate. By default, a signed OpenSSH certificate stays valid forever. You must set an expiration time so the agent loses access after the task. The certificate can also hold options that limit what the session can do.
What native SSH setup solves and what it misses
Native OpenSSH certificates solve the problem of static keys sitting on agent disks. You do not have to copy public keys to every server ahead of time. The target server trusts your certificate authority key and checks the principal name. When the time window ends, the certificate stops working without manual cleanup. Using a bastion or jump host also gives your team a single entry point.
Still, host login approval only controls who gets through the front door. Once the SSH session starts, the server trusts every command that user role runs. An approved agent might still run a bad script or drop a vital service. Native SSH login checks do not ask a human to review each shell command. They also do not rewrite sensitive data like emails or phone numbers in outputs. If an agent reads a customer file, raw card numbers or national IDs appear. Masking tools must rewrite those matched fields in results before the agent reads them.
Approval Chains & Escalation + SSH Access Management: architecture patterns and best practices
Continue reading? Get the full guide.
Free. No spam. Unsubscribe anytime.
Why per-command checks matter for AI agents
An AI agent works fast and runs many shell commands in a few seconds. Approving a whole SSH session gives the agent a blank check for that window. Even a five-minute certificate gives an agent enough time to delete a folder. Teams need a way to check individual commands while the session is live.
Splitting read actions from write actions makes human review much easier for your team. You can let the agent check disk space or read app logs right away. When the agent tries to restart a service, the system pauses for human approval. This keeps the agent helpful while stopping bad changes on your live servers.
Check SSH commands and file transfers on the host
OpenSSH keys and certificates decide who can log into a Linux host. They do not check what a one-shot command or file transfer does once the session starts. With hoop.dev, you run an open-source SSH server with guardrails on the host. It checks callers with OpenSSH certificates and works with ssh, sftp, scp, and rsync clients.
When a caller runs a one-shot command or moves a file, hoop.dev guardrails check the command or path before it runs. Then hoop.dev data masking masks matched sensitive patterns in downloads and output. Interactive shells carry no guardrails because keystrokes have no statement boundary. Teams can turn interactive shells off and allow only commands and file tasks. The audit trail logs who connected and what was allowed or refused, without storing keystrokes or file bytes. See the source on GitHub.
Frequently asked questions
Can I use static SSH keys for an AI agent?
You can, but static keys stay active until someone removes them from the server. Signed OpenSSH certificates with short time limits are much safer for agent tasks.
How do OpenSSH certificates limit which server an agent uses?
When you sign a key, you list allowed user or host names as principals. The target server checks this principal list before it lets the agent log in.
Does approving an SSH login stop an agent from reading secrets?
No, login approval only starts the shell session for that Linux user role. Any file or secret that user can read is open to the agent.
Related guides
In this post