title: "Tuskira open-sources AI Agent Gateway for enterprise control" slug: "tuskira-open-sources-ai-agent-gateway-for-enterprise-control" published: "2026-10-11" beat: "Tools" tags: ["Tools", "Launches"] creator: "Agentry Newsroom" editor: "Susanne Sperling, Editor — Human in the Loop" tools: ["Claude (Anthropic)", "Perplexity Sonar"] creativeWorkStatus: "verified" dateReviewed: "2026-10-11" aiActArticle50: "compliant" humanView: "https://agentry.news/launches/tuskira-open-sources-ai-agent-gateway-for-enterprise-control" agentView: "https://agentry.news/agent/tuskira-open-sources-ai-agent-gateway-for-enterprise-control"
Tuskira released an open-source, self-hosted AI Agent Gateway on October 1, 2026, giving developers, platform teams, and security teams direct control over agent infrastructure, model switching, polic
Drafted by an AI agent. Verified by Susanne Sperling, Editor — Human in the Loop. AI policy.
Tuskira released the Tuskira AI Agent Gateway as an open-source project on October 1, 2026, according to a Business Wire press release dated San Francisco. The gateway, published on GitHub under the Apache 2.0 license, is a self-hosted, open-source infrastructure layer that sits between AI agents and the language-model providers, MCP servers, and tools they consume.
The gateway enables developers, platform teams, and security teams to inspect and operate agent infrastructure from within their own environments—without requiring a Tuskira account or reliance on a hosted service. According to Tuskira's announcement, the gateway can configure models and tools, apply policy, control access, log calls, track token spending, and switch models without reconfiguring agents. This architecture addresses a core friction point in production agent deployment: the need to observe and govern agentic behavior across multiple LLM providers and tool integrations without rebuilding agent code each time infrastructure changes.
Tuskira released the gateway as pre-1.0 alpha software, currently at v0.4.0. The GitHub repository contains source code, documentation, and CI/CD workflows, signaling that the project is actively developed and open to community contribution. The gateway's placement in the agent stack—between decision-making logic and external services—means it can enforce policy, meter usage, and route requests without agents needing to know about infrastructure changes downstream.
The release reflects growing friction in enterprise agent adoption. As organizations move beyond single-agent prototypes into multi-agent systems spanning different LLM providers (OpenAI, Anthropic, open-source models) and tool ecosystems, the operational overhead of managing governance, compliance, and model switching has become acute. A self-hosted gateway that abstracts this layer lets security teams apply organization-wide controls without forcing engineering teams to rewrite agent logic.
Source code is available on GitHub under the Tuskira organization, with full observability documentation for operators. The Apache 2.0 license permits commercial use, modification, and distribution, lowering adoption friction for enterprises that require license flexibility.