<feed xmlns="http://www.w3.org/2005/Atom"> <id>https://blog.christianposta.com/</id><title>Christian Posta</title><subtitle>Christian Posta's blog on Istio, Envoy, API gateways, MCP, and the infrastructure that runs production AI agents.</subtitle> <updated>2026-10-08T20:23:46+00:00</updated> <author> <name>Christian Posta</name> <uri>https://blog.christianposta.com/</uri> </author><link rel="self" type="application/atom+xml" href="https://blog.christianposta.com/feed.xml"/><link rel="alternate" type="text/html" hreflang="en" href="https://blog.christianposta.com/"/> <generator uri="https://jekyllrb.com/" version="4.4.1">Jekyll</generator> <rights> © 2026 Christian Posta </rights> <icon>/assets/img/favicons/favicon.ico</icon> <logo>/assets/img/favicons/favicon-96x96.png</logo> <entry><title>Human-in-the-loop Authorization Patterns for Agents</title><link href="https://blog.christianposta.com/human-in-the-loop-authorization-patterns-for-agents/" rel="alternate" type="text/html" title="Human-in-the-loop Authorization Patterns for Agents" /><published>2026-08-24T23:17:52+00:00</published> <updated>2026-10-08T20:22:24+00:00</updated> <id>https://blog.christianposta.com/human-in-the-loop-authorization-patterns-for-agents/</id> <content type="text/html" src="https://blog.christianposta.com/human-in-the-loop-authorization-patterns-for-agents/" /> <author> <name>Christian Posta</name> </author> <summary>In my work with customers adopting and securing AI agents at scale in enterprise environments, we are running into “human in the loop” usecases. For example, usecases where a sensitive operation MUST be approved by a human; sometimes even a different human than the one on whose behalf the agent is acting. As part of my work on Agent Identity and Access Management here at Solo.io, I research var...</summary> </entry> <entry><title>APIs for Probabilistic Callers</title><link href="https://blog.christianposta.com/apis-for-probabilistic-callers/" rel="alternate" type="text/html" title="APIs for Probabilistic Callers" /><published>2026-08-11T21:34:28+00:00</published> <updated>2026-08-11T21:34:28+00:00</updated> <id>https://blog.christianposta.com/apis-for-probabilistic-callers/</id> <content type="text/html" src="https://blog.christianposta.com/apis-for-probabilistic-callers/" /> <author> <name>Christian Posta</name> </author> <category term="AI Agents" /> <category term="Identity" /> <summary>APIs designed for AI agents need to enforce resource-owned invariants or surface them into the enforcement path, not just expose operations guarded by caller-side authorization checks.</summary> </entry> <entry><title>Credential Brokering Patterns for AI Agents Part 2: Protecting the Token Vault with AWS KMS</title><link href="https://blog.christianposta.com/credential-brokering-patterns-for-ai-agent-egress-part2/" rel="alternate" type="text/html" title="Credential Brokering Patterns for AI Agents Part 2: Protecting the Token Vault with AWS KMS" /><published>2026-08-03T16:00:00+00:00</published> <updated>2026-08-03T16:00:00+00:00</updated> <id>https://blog.christianposta.com/credential-brokering-patterns-for-ai-agent-egress-part2/</id> <content type="text/html" src="https://blog.christianposta.com/credential-brokering-patterns-for-ai-agent-egress-part2/" /> <author> <name>Christian Posta</name> </author> <category term="AI Agents" /> <category term="Identity" /> <summary>A credential broker fixes the agent's credential problem by concentrating credentials somewhere else. Here's how we keep that concentration from becoming a single point of total failure.</summary> </entry> <entry><title>Credential Brokering Patterns for AI Agents Part 1: Don't Give the Agent the Keys</title><link href="https://blog.christianposta.com/credential-brokering-patterns-for-ai-agent-egress/" rel="alternate" type="text/html" title="Credential Brokering Patterns for AI Agents Part 1: Don&amp;apos;t Give the Agent the Keys" /><published>2026-07-27T10:41:23+00:00</published> <updated>2026-08-04T04:26:51+00:00</updated> <id>https://blog.christianposta.com/credential-brokering-patterns-for-ai-agent-egress/</id> <content type="text/html" src="https://blog.christianposta.com/credential-brokering-patterns-for-ai-agent-egress/" /> <author> <name>Christian Posta</name> </author> <category term="AI Agents" /> <category term="Identity" /> <summary>The safest credential for an AI agent is one it never holds. A look at the CB4A draft, why the proxy model is the one that actually works against today's bearer-token APIs, and how we build it in agentgateway.</summary> </entry> <entry><title>Okta SAML and Keycloak for ID-JAG Cross App Access</title><link href="https://blog.christianposta.com/okta-saml-and-keycloak-for-id-jag-cross-app-access/" rel="alternate" type="text/html" title="Okta SAML and Keycloak for ID-JAG Cross App Access" /><published>2026-07-13T01:05:21+00:00</published> <updated>2026-07-15T22:59:25+00:00</updated> <id>https://blog.christianposta.com/okta-saml-and-keycloak-for-id-jag-cross-app-access/</id> <content type="text/html" src="https://blog.christianposta.com/okta-saml-and-keycloak-for-id-jag-cross-app-access/" /> <author> <name>Christian Posta</name> </author> <category term="AI Agents" /> <category term="Identity" /> <summary>A working Okta SAML/SSO → ID-JAG → Keycloak access-token path for Cross-App Access</summary> </entry> </feed>
