Harness-Benefit Gain (Δbenefit)

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• Harness-Updating Gain (Δupdate)method0.847
  Metric measuring harness-updating capability as the mean pairwise gain across an anchor agent set
• Harness-Benefit Capabilityconcept0.801
  The capability of a task-solving agent to benefit from updated harnesses during task solving
• Harness-benefit is non-monotonic in base capability: weak-tier models benefit little, mid-tier models benefit most, and strong-tier models benefit less than mid-tierclaim0.754
  Second major claim of the paper, supported by Δbenefit measurements across six models on three benchmarks
• On MCP-Atlas, harness-benefit peaks at GPT-OSS-120B (7.0 pp), with lower gains at both ends of the base-capability scalefinding0.708
  Replication of non-monotonic harness-benefit pattern on a second benchmark
• On SWE-bench, harness-benefit peaks at Qwen3-235B (19.3 pp), while weaker Qwen3-32B gains only 4.4 pp and stronger Opus 4.6 gains only 2.6 ppfinding0.704
  Core finding demonstrating non-monotonic relationship between base capability and harness-benefit
• Agent Harnessconcept0.698
  The external non-parametric context and infrastructure (prompts, skills, memories, tools) through which an LLM is deployed for task execution
• which models actually benefit from updated harnesses?question0.695
  Second open question the paper sets out to answer through agent-side analysis
• Qwen3.5-9B evolver achieves highest harness-updating gain on SkillsBench (3.8 pp), exceeding Claude Opus 4.6 (2.3 pp) and Qwen3-235B (1.5 pp)finding0.694
  Case demonstrating that model scale does not predict harness-updating quality