Unverified50% confidenceFactExact time
Harness2's working loop is: the Agent proposes modifications to the existing Harness for a new task, runs several candidate versions in parallel, and selects the better one by comparing actual performance.
1
Sources
50%
Confidence
Long-term
Relevance
10/19/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe core idea of Harness2 is to recursively rewrite the agent harness at test time without updating the model weights.79% similarVerifiedThe concept of Harness originates from Test Harness in software engineering, and in the AI Agent context it is extended to mean the systematic engineering that enables an Agent to run continuously, cyclically, and in a controlled manner75% similarUnverifiedThe harness is the intermediate layer in an Agent architecture responsible for orchestrating large model inference and external tool calls, and its design complexity directly affects the overall performance of the Agent.73% similarUnverifiedHarness breaks a single agent job into Turns and Steps, where one Turn can contain multiple Steps, and each Step initiates a model request, receives output as a stream, and handles tool calls and results72% similarUnverifiedHarness Engineering的核心是让Agent知道目标边界、状态追踪、自我修正和恢复机制71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1064756API
curl https://kongchang.com/api/v1/knowledge/claims/1064756MCP
get_claim(id=1064756)