Noma Security

No entries match these filters. Clear filters

1. announced Sep 28, 2026 · issue 089

   Noma extends its agent control plane to employee endpoints, discovering coding and productivity agents via existing EDREDREndpoint detection and response, monitoring computers for suspicious behavior and providing investigation and response controls, such as isolating a device./MDMMDMMobile device management, centrally configuring and managing enrolled devices. Administrators use it to apply settings, distribute applications and enforce device security policies. infrastructure.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

   • Extends Noma's agent inventory to employee endpoints, using existing EDR or MDM deployments to discover agents, MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. servers, skills, connectors, and active processes on managed devices.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
   • Adds a governed registry that automatically assigns every discovered agent, MCP server, and skill a status of approved, needs review, or blocked.
   • Adds identity-aware Access Control policies that attribute each endpoint agent to the human behind it, linking to IdPIdPA service that authenticates users and supplies identity information to other applications. It is the sign-in authority behind many single sign-on systems. users and groups for per-user, per-group, or org-wide enforcement.
   • Adds tool- and action-level Access Control granularity, allowing read access to be broadly permitted while create, update, or delete operations are restricted to specific groups.
   • Adds AI-DR runtime threat protection for endpoint agents, detecting prompt injectionprompt injectionAn attack where text the model reads, such as a web page, file or tool result, carries instructions that override the user's. It is the main security risk for agents with tools. (including encoded, obfuscated, and indirect injection via tool responses), sensitive data leakage (PIIPIIPersonally Identifiable Information — data that can identify a specific individual, such as names, email addresses, or SSNs. Cyber tools flag or redact it to limit exposure and meet privacy regulations., PCIPCIA hardware bus standard that connects peripheral devices to a computer's motherboard; cyber tools targeting hardware-level forensics or memory analysis use it to access device registers and DMA regions directly., secrets), malicious intent, tool poisoning, scope violations, and behavioral drift.

   +4 more show less

   • Adds Contextual Policies within AI-DR that correlate prompts, tool calls, tool responses, and reasoning across full sessions to surface multi-step attacks such as indirect prompt injectionindirect prompt injectionAn attack where text the model reads, such as a web page, file or tool result, carries instructions that override the user's. It is the main security risk for agents with tools. followed by secret exfiltration.
   • Adds Adaptive Context Retrieval, which learns an application's traffic patterns to distinguish reusable context (system prompts, session history, RAGRAGRetrieval-augmented generation: looking up relevant documents and adding them to the prompt before the model answers. It grounds responses in private or current data without retraining. context) from user-controlled input, improving detector signal and reducing false positives.
   • Extends Open Enforcement to the endpoint, enabling the same Access Control, Agent Boundaries, and AI-DR policies to be enforced through agent hooks, AI and MCP gateways, SDKs, direct APIs, and existing EDR/MDM infrastructure.
   • Adds coverage for coding and productivity agents including Claude Code, Claude Cowork, Cursor, Codex, Windsurf, Kiro, Antigravity, and OpenClaw.
   • Extends Noma's agent inventory to employee endpoints, using existing EDR or MDM deployments to discover agents, MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. servers, skills, connectors, and active processes on managed devices.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
   • Adds a governed registry that automatically assigns every discovered agent, MCP server, and skill a status of approved, needs review, or blocked.
   • Adds Contextual Policies within AI-DR that correlate prompts, tool calls, tool responses, and reasoning across full sessions to surface multi-step attacks such as indirect prompt injectionindirect prompt injectionAn attack where text the model reads, such as a web page, file or tool result, carries instructions that override the user's. It is the main security risk for agents with tools. followed by secret exfiltration.
   • Adds Adaptive Context Retrieval, which learns an application's traffic patterns to distinguish reusable context (system prompts, session history, RAGRAGRetrieval-augmented generation: looking up relevant documents and adding them to the prompt before the model answers. It grounds responses in private or current data without retraining. context) from user-controlled input, improving detector signal and reducing false positives.

   5 more in Everything