Runlayer · Security requirement
Agent-to-agent communication security
Research finding
Runlayer materials reviewed did not provide a public claim for trust or policy enforcement between agents.
Public-source research
Vendor evidence
Review what vendors say publicly, the exact quoted source text, the security requirement each statement may support, and what still needs verification.
Research library coverageCounts describe the research workflow, not vendor quality or product effectiveness.View details
Evidence records1280
Source checked863
Needs verification0
Source captured0
No supporting claim found411
Excluded from evidence6
Showing 261–280 of 1280 evidence records
Runlayer · Security requirement
Non-human identity and service-account security
Research finding
Runlayer materials reviewed did not provide a public claim for non-human identity, service-account, secret, or workload credential lifecycle controls.
Runlayer · Security requirement
AI agent identity and permissions
What the vendor says
Runlayer claims identity-aware access policy and scoped permissions for users and agents.
Exact source quote
under which identity, budget, OAuth grant, and runtime conditions
Runlayer · Security requirement
AI coding-agent and workstation security
What the vendor says
Runlayer claims centralized controls for employees using Claude, Cursor, ChatGPT, Codex, internal agents, Model Context Protocol (MCP) servers, and existing AI clients.
Exact source quote
employees are adopting Claude, Cursor, ChatGPT, Codex, and internal agents
Keycard · Security requirement
Unapproved AI use discovery
Research finding
Keycard materials reviewed did not provide a public claim for enterprise-wide discovery of unapproved AI use.
Keycard · Security requirement
AI-feature discovery in business applications
Research finding
Keycard materials reviewed did not provide a public claim for embedded AI discovery across the business-application environment.
Keycard · Security requirement
Approved AI usage monitoring
What the vendor says
Keycard claims visibility and attribution for agent actions, tool calls, policy decisions, and data access.
Exact source quote
See every action your agents take. Know exactly who authorized it.
Keycard · Security requirement
Controls for unapproved AI use
Research finding
Keycard materials reviewed did not provide a public claim for policy enforcement over unapproved AI use.
Keycard · Security requirement
Sensitive-data protection for generative AI
Research finding
Keycard materials reviewed did not provide a public claim for inspection or enforcement over sensitive data moving through AI.
Keycard · Security requirement
Browser and business-application controls
Research finding
Keycard materials reviewed did not provide a public claim for browser or software as a service (SaaS)-session controls for employee AI use.
Keycard · Security requirement
Generative AI application security
Research finding
Keycard materials reviewed did not provide a public claim for security controls for custom generative AI applications at runtime.
Keycard · Security requirement
AI governance, risk, and compliance
Research finding
Keycard materials reviewed did not provide a public claim for enterprise AI governance, risk, approval, and compliance workflows.
Keycard · Security requirement
AI assurance and adversarial testing
Research finding
Keycard materials reviewed did not provide a public claim for adversarial testing or release assurance for AI systems.
Keycard · Security requirement
AI model and supply-chain security
Research finding
Keycard materials reviewed did not provide a public claim for model and AI component supply-chain inspection.
Keycard · Security requirement
AI gateway, tool-connection, and runtime controls
What the vendor says
Keycard claims authentication and task-scoped access controls for Model Context Protocol (MCP), CLI, application programming interface (API), and downstream services.
Exact source quote
Add auth to any agent surface - MCP, CLI, or API.
Keycard · Security requirement
Action-taking agent monitoring
What the vendor says
Keycard claims a real-time event stream of agent actions, tool calls, policy decisions, and attributed audit events.
Exact source quote
A real-time event stream of every agent action, tool call, and policy decision.
Keycard · Security requirement
Agent-to-agent communication security
What the vendor says
Keycard claims authentication for agent-to-agent delegation while preserving user identity.
Exact source quote
Announcing Keycard for Multi-Agent Apps
Keycard · Security requirement
Non-human identity and service-account security
What the vendor says
Keycard claims workload attestation and short-lived credentials that avoid long-lived secrets and over-permissioned service accounts.
Exact source quote
No long-lived secrets, no over-permissioned service accounts.
Keycard · Security requirement
AI agent identity and permissions
What the vendor says
Keycard claims composite user, device, agent, and task identity with runtime policy and task-scoped credentials.
Exact source quote
Identity = user + device + agent + task
Keycard · Security requirement
AI coding-agent and workstation security
What the vendor says
Keycard claims governance of coding-agent shell, script, Model Context Protocol (MCP), environment, credential, and audit paths.
Exact source quote
keycard run virtualizes .env and mcp.json so credentials never hit disk, and every execution path is audited.