Runlayer Agents
Background & hosted agents
Platform overview
One platform to accelerate and control AI usage
Runlayer gives teams a control plane for approved AI usage, a gateway for MCP requests, and the Runlayer Agents runtime for agents created on Runlayer infrastructure.
How Runlayer works
Runlayer gives employees approved AI capabilities while AI Platform, AI Transformation, and IT & Security control what can be used, who can use it, and what happened.
Enablement
Runlayer Agents
On-demand agents running on Runlayer infrastructure. Employees can create agents for recurring work while platform owners keep approved tools, scoped credentials, policy, runtime security, and activity history in place.
Runlayer Catalog
Reusable skills and plugins teams can publish and share with agents. The catalog gives useful AI capabilities an owner, approval state, dependency context, and usage history.
MCP Gateway
Approved MCP access across AI clients and third-party agents. MCP requests route through the gateway so identity, policy, runtime security, and audit stay attached before company systems are reached.
Control
Shadow AI
Shadow AI discovery and response. Watch finds shadow MCPs, skills, plugins, and client configs so teams can migrate, approve, remove, or block the right findings.
Runtime Security
Execution-time security checks for AI tool use. Runtime Security inspects risky tool behavior across the gateway and agentic runtime before sensitive actions continue.
Agent IAM & Governance
Actor-aware access control for AI usage. Runlayer evaluates who is acting, which client or agent is involved, what tool or resource is requested, and which conditions apply.
ROI & Observability
Adoption, decisions, findings, and audit history. ROI & Observability shows what teams use, where access is blocked, which findings matter, and what happened during agent activity.
Built for enterprise security and deployment
The platform fits into the systems enterprises already use to manage identity, devices, infrastructure, and audit evidence. Keep tamper-proof logs, security events, and usage history available for review.
SSO and SCIM
Use existing identity groups for provisioning, policy, and access review.
Policy model
Map access to users, groups, roles, attributes, agent accounts, clients, tools, resources, and runtime conditions.
Self-hosted deployment
Run everything in your own infrastructure.
MDM rollout
Use existing device management to deploy approved client configs, shadow usage detection, and enforcement.