As a founder in ForgeRock, I have had the privilege to watch and participate in the digital identity evolution. While significant progress has been made in how we authenticate identities and manage their access, most of this has focused on human identities.
A few years ago, I saw a need to rethink the whole model of identity, realizing it is much more than humans. In fact, it is much more than the bounds of the identity market as well. The whole way that we address how identities interact felt too static, too siloed, too handcuffed. It also lacked the perspective of how data is used at an organization.
I founded IndyKite with the vision that identity is one part of the bigger machine operating in the enterprise - one that moves beyond humans to devices and things, one that requires intelligence and operating context. One that absolutely must take data management and operational data use into account.
And there has been no other time in my career where I have felt and seen the excitement for how this model is perfectly served for the Agentic AI era.
As we unleash agents into our businesses to help us accelerate work, some of the same challenges exist but some wholly new ones as well. Contextual decisions such as authorization, workflow execution, escalation to human review, policy enforcement, and dynamic data handling must happen at runtime. As these decisions become increasingly autonomous and interconnected, the importance of trust, provenance, auditability, and traceability becomes more critical than ever.
This is a new era and it requires a new architectural model and a new way of thinking.
A key part of that is how to set up agents to succeed.
AI agents often lack the operational context behind the platforms they are building against. This can result in architectural, governance or operational failures that only emerge in production.
To help close that gap for developers building with IndyKite, we’ve launched IndyKite skills on skills.sh - an open registry for reusable AI agent skills.
We’ve open sourced IndyKite skills that developers can install directly into tools like Claude Code, Cursor, Codex, and other coding agents. The skills provide agents with platform-specific implementation guidance, architectural patterns, workflows, and operational knowledge during development.
IndyKite skills help agents work more effectively with ContX IQ, AgentControl, runtime policy enforcement, trust scoring, delegated execution chains, and policy-aware data interactions.
An agent may know how to call an API, however it may not understand how runtime enforcement should be structured, how graph-driven retrieval patterns operate, or how trust and policy evaluation should function in production systems.
This becomes increasingly important as AI agents move deeper into enterprise software development and operational environments.
IndyKite was built around the idea that context allows systems to make better decisions and operate smarter. That is ultimately what these skills are about. Leveraging context to help developers and agents build autonomous systems that behave correctly in real enterprise environments.
Excited to contribute to the ecosystem and looking forward to expanding IndyKite skills as we build!
