NEW RUNTIME

Topic hub

Harness Engineering

A New Runtime topic hub collecting signals, patterns, field notes, and public sources about harness engineering.

Short answer

  • Harness Engineering is tracked here as an evidence-linked topic, not as a static glossary entry.
  • The page connects raw observations to pattern hypotheses, longer analysis, and public sources.
  • Use it as the canonical landing page before drilling into individual records.

Pattern memory

Current hypotheses

1 patterns
  1. high

    Harness architecture outlives model choice

    For production agents, the harness is becoming a more durable product boundary than the identity of the model running inside it.

Raw signals

Recent observations

5 signals

  1. OpenCode makes the model a swappable dependency

    OpenCode keeps rules, skills, permissions, sessions, and tools stable while the underlying model changes.

  2. A local agent is a runtime, not one model

    A local agent stack combines models, orchestration, memory, skills, MCP tools, permissions, judges, and output guards.

  3. Deep Agents packages a harness for long-running tasks

    LangChain's Deep Agents combines planning, files, memory, subagents, middleware, and backends for multi-step work.

  4. Deep Agents combines plans, files, memory, and subagents

    The open-source Deep Agents stack packages common long-horizon primitives on top of LangGraph and LangChain.

  5. Claude Code Can Build a Task-Specific Harness Before Coding

    Instead of applying one universal setup, the agent creates targeted scripts, checks, and context for the current task before entering the implementation loop.