Hue

Hue is a testing and observability platform for teams building AI agents. It turns real user requests into repeatable tests, so an agent can be rerun against the same context and data it had in the original request before a change ships. The product has four parts. Observability traces real requests, model calls and tool activity. Eval Suggestions finds tasks worth replaying from real usage and highlights coverage gaps. Agent Replays rerun a task in a stateful environment, where each run gets its own world that keeps every change and can be reset or customized. Evaluations score each rerun and compare performance across agent versions and models. Tools in these environments are simulated APIs and MCP servers that the vendor says match the real ones, including edge cases, with examples such as Slack, Salesforce, Notion, Gmail and Microsoft 365. It is built for developers shipping agents. Instrumentation works through Hue's own SDK, which supports most languages, or through OpenTelemetry traces from any compatible tool. The vendor lists SOC 2 Type II, ISO 27001 and HIPAA compliance, project scoped keys and controls over what traces and context are captured. Sign up is self serve, and no pricing is published.