Pulumi Launches Context API to Unify Infrastructure State as an Agent-Ready Graph

Pulumi has launched the public preview of the Pulumi Context API for Enterprise and Business Critical customers, establishing a unified, queryable graph across an organization's entire cloud landscape. The engine connects Pulumi-managed resources, stack-to-stack dependencies and outputs, and unmanaged cloud resources indexed through Pulumi's discovery capabilities. Interfacing with the graph is handled via JSON query documents featuring anchor, traverse, and return semantics, accessible over REST endpoints or the Pulumi CLI using the `pulumi api GraphQuery` command in CLI version 3.243.0 and newer. Crucially, the endpoint exposes a live Markdown schema primer that enables AI agents to self-onboard and self-correct queries automatically. For platform engineers and SREs, this marks a substantial evolution from flat, siloed state files toward relational infrastructure intelligence. In enterprise cloud deployments, IaC architectures are deliberately split into modular micro-stacks to limit blast radius; however, this architecture historically obscures transitive dependencies. Engineers and autonomous agents operating on an isolated stack frequently lack immediate visibility into whether shared virtual networks, routing tables, or IAM policies impact adjacent services. By unifying declared state with discovered cloud assets in an indexed graph, the Context API provides a reliable mechanism to assess systemic impact prior to provisioning. This release reflects a broader paradigm shift across DevOps: bridging declarative Infrastructure as Code with the emerging demands of agentic automation. As coding agents transition from generating boilerplate code to proposing direct production mutations, providing deterministic ground truth is vital. Traditional language models operating without real-time graph context are susceptible to generating drift or breaking unreferenced dependencies. Exposing structured graph queries alongside machine-readable schema contracts establishes the guardrails necessary for autonomous tooling to reason safely over enterprise architecture. In practice, platform teams should evaluate integrating the Context API into pull request validation pipelines and internal developer platforms to automate dependency and blast-radius checks. While the current preview focuses on infrastructure resources and stack linkages, Pulumi plans to incorporate Pulumi ESC configuration environments, team RBAC hierarchies, and IDP service catalogs into the graph vocabulary over time. Engineering leads deploying agentic workflows should wire the graph schema endpoint directly into their agents' toolsets to ensure AI-driven changes are validated against live topological data before executing deployment pipelines.