AgentGateway
PulseAugur coverage of AgentGateway — every cluster mentioning AgentGateway across labs, papers, and developer communities, ranked by signal.
- 2026-08-01 product_launch AgentGateway released version 1.4.0, fixing a security vulnerability. source
7 day(s) with sentiment data
-
AI research covers market making and ambivalence; enterprise tool discovery explored
Two research papers explore distinct AI applications: one details a market-making game with Nash equilibria and learning strategies, while another introduces a multimodal framework for recognizing ambivalence and hesita…
-
Solo Enterprise adds LLM cost management with virtual keys and budgets
The Solo Enterprise AgentGateway now offers cost management features, allowing teams to track and cap their LLM spending. This system assigns a price to each LLM request, attributes it to a specific user and team, and e…
-
Solo.io agentgateway vulnerability allows sensitive file access
A security vulnerability has been identified in Solo.io's agentgateway where a filesystem MCP server, even when scoped to a specific directory, can still access sensitive files like .env if an agent requests them by pat…
-
Open-weight LLMs tested as cheaper AI agent tiers against Claude Sonnet 5
An experiment tested open-weight language models as a cost-effective alternative to proprietary models for AI agents. Using agentgateway, two locally hosted open-weight models, Qwen3.8 and Nemotron-3.5-Lightning, were c…
-
MCP Gateway Centralizes Agent Tool Security
An MCP gateway serves as a centralized security layer for agents interacting with internal MCP servers, addressing the challenge of managing tool access and permissions across multiple agents and teams. Unlike embedding…
-
AI tool descriptions exploited for data theft; gateway rules prove effective
Security researchers have identified a vulnerability where malicious instructions can be embedded within tool descriptions for AI models, a technique termed "tool poisoning." A specific instance involved the `postmark-m…
-
agentevals shifts to OpenTelemetry for AI evaluation data
The agentevals open-source project has released version 0.10.0, shifting its evaluation ingestion and output to use OpenTelemetry (OTel) standards. This update replaces its previous WebSocket-based ingestion with OTLP (…
-
MCP Cross App Access Explained: Enterprise IdP Controls Server Access
The MCP Cross App Access (XAA) extension enables enterprise identity providers to manage user access to MCP servers. This process involves a three-step HTTP exchange where a client obtains an ID token from the IdP, exch…
-
MCP spec revision enables stateless Kubernetes operations, boosting scalability
A new revision to the MCP specification, released on July 28, 2026, enables stateless operation for MCP servers on Kubernetes. This update eliminates the need for session affinity, which previously caused scaling and re…
-
AgentGateway deployed on SAP BTP Kyma for secure MCP gateway setup
This guide details the setup of AgentGateway on SAP BTP Kyma for production-grade MCP gateway deployment. It covers essential security and observability features such as JWT authentication via XSUAA, CEL tool authorizat…
-
MCP Gateways Emerge as Key Infrastructure for Enterprise AI Integration
The Model Context Protocol (MCP) is emerging as a critical standard for securely integrating AI models with enterprise systems, addressing the 'plumbing' challenges that often hinder AI project success. Several platform…
-
Rate Limiting Essential for MCP Servers to Prevent Overload and High API Costs
This blog post explains the importance of rate limiting for MCP (Model Communication Protocol) servers to prevent issues like denial-of-service attacks, excessive memory consumption, and high API bills. It details how r…
-
AI Gateways Emerge as System Integration Evolves
A curated list of system integration tools and patterns, initially created in 2021, has evolved significantly over five years. The most notable change is the emergence of an AI gateway layer, which addresses familiar AP…
-
AgentGateway config flaw mirrors football rule; Langflow hit by critical CVE
An advisory from AgentGateway revealed a security vulnerability in its configuration, specifically affecting stateful setups with multiple backends. The issue, present before version 1.4.0, allowed callers to bypass aut…
-
Verifying Multi-Cluster MCP Paths for Agent Execution in Kubernetes
This technical post outlines a verification process for multi-cluster MCP (Multi-Cluster Protocol) paths, essential for ensuring agents can reliably execute tasks across different Kubernetes clusters. It details a step-…