Robokeys
Please note that as of March 2026, the commercial entity Robokeys Limited has been disolved.
This administrative change does not affect the availability or development of the open-source projects. Ownership and maintenance of the code repositories (including the Robokeys platform) have been transferred to Rob Deas, who will continue to lead development and support the community.
Commercial support and consulting remains available directly through Rob Deas. s.
AI-Assisted SSH Security for the Age of Automation
In a world where AI agents are everywhere, SSH security needs a rethink. ShellGuard delivers that rethink.
Commands submitted via WebSocket API
AI analyzes potential impact and risk level
High-risk commands require manual approval
Approved commands executed with full audit trail
Real-time dashboard for monitoring and approving AI-requested terminal actions. See pending commands, risk assessments, and approve or reject with full context.
Perfect for developers and AI system integration. Test commands manually or script interactions using the same WebSocket interface that AI systems use.
Automation is accelerating rapidly, but so is the blast radius of mistakes. AI agents offer immense power, but a single unsupervised command can be catastrophic.
Commands are automatically scored for potential risk using AI and configurable rules before execution. Never worry about destructive commands slipping through.
High-risk commands are routed to an Approval Center where authorized personnel can approve or reject them with full context and risk analysis.
Every command, its risk score, approval decision, and execution result is logged for full accountability and compliance requirements.
Modern web UI built with HTMX provides live updates on command pipelines, approvals, and system status without page refreshes.
Easily integrate custom risk assessors and approval logic to match your security policies and organizational requirements.
Engineered with Kotlin and Spring Boot for reliability, performance, and enterprise-grade deployment scenarios.
This is the actual production interface for the AI Command Approval Centre - click the picture to expand
ShellGuard fits perfectly into our broader mission to build secure and intelligently controlled automation frameworks.
Semantic metadata standard for AI-assisted development
Versioned safe experimentation without polluting Git history
🔍 Want the Technical Deep Dive?
Read our detailed technical documentation and implementation guides
More comming soon !
📖 Download User Guide (PDF)ShellGuard is open source and ready to deploy. Join the community building the future of secure AI operations.
We welcome feedback, contributions, and ideas to make this even more impactful.