Home / UHP Glossary: Agent Harness, Session, Response, Conformance

Reference

UHP glossary

Short definitions for the terms used across the Unified Harness Protocol specification and this independent guide.

Verified: 15 Aug 2026Protocol: 2026-08-11No executable JS

Core UHP terms

Agent harness: A complete agent runtime around a model, including its execution loop, tools and session state.
Base: A stable identifier for a harness family, such as codex, claude-code or hermes.
Configured harness: A harness base plus runtime configuration such as model defaults, instructions, restrictions, skills, MCP servers and budgets.
Client: The application speaking UHP to request work.
Server: The UHP implementation that accepts tasks and drives one or more harnesses.
Response: One atomic unit of work: a task request and its resulting state/output.
Session: A chain of responses sharing conversation context, workspace and configured harness.
Container: The file namespace associated with a session.
Artifact: A file produced by agent work and made available through the protocol.
Event: A streamed fact describing progress, delivered through the event stream.
Conformance class: Core, Extended or Full; each class defines a cumulative set of mandatory server capabilities.
Capability discovery: The server document at GET /v1/uhp describing versions, conformance class and supported features.

MCP: Model Context Protocol, an open protocol for connecting AI applications to tools, resources and prompts. It addresses a different boundary from UHP.
Harness Protocol: A separate project for portable harness configuration using harness.yaml.
AHP: Agent Harness Protocol, a separate project oriented around agent/harness supervision events and policy decisions.