DeltaHarness: Self-Evolving Harnesses for Resource-Adaptive Tool Execution
Organizations: University of Science and Technology of China · Beijing Normal University · Zhejiang University · Dmall Inc. ·
Abstract
Tool-using agents rely on execution harnesses to organize and schedule tool operations. The same task and resource description can conceal runtime conditions that reverse the latency ordering of valid execution plans. Resolving this uncertainty requires firsthand execution in the target setting. We introduce DeltaHarness, an execution-guided policy-search method that adapts agent harnesses from cold-start measurements. Its QHI loop uses the observed response to update runtime dispatch, revise physical plans, and select experiments for later proposals. An independent confirmation panel controls policy replacement. With fixed dispatch, a revised materialization policy reduces complete-delivery latency by 30.0% on one video-editing task across three fresh pairs. On held-out tasks, the evolved policies reduce latency by 8.2% relative to the best fixed plan in one seed and by 4.6% relative to the initial proposal in another.
Figures & tables
| Trial | Order | (s) | (s) | Ratio |
|---|---|---|---|---|
| 1 | 9.74 | 14.06 | 0.693 | |
| 2 | 9.75 | 14.27 | 0.683 | |
| 3 | 9.73 | 13.42 | 0.725 | |
| Paired geometric mean | 0.700 | |||
| Comparison | Ratio | 95% CI |
|---|---|---|
| S / Uniform | 0.984 | [0.966, 1.001] |
| S / Balanced | 1.173 | [1.103, 1.251] |
Appendix figures & tables11 assets
Supplementary material from the paper’s appendix.
Appendix
| Parameter | Physical choice |
|---|---|
| Number of output ranges, 1–32 | |
| Range pool | Concurrent range operations |
| Cut mode | Even or clip-aware cuts |
| LPT | Order ranges by estimated work |
| Hardware decode | Request when supported |
| Materialize | Create intermediate clips |
| Seed A | Seed B | |||||
|---|---|---|---|---|---|---|
| Policy | Ratio | 95% CI | Unseen | Ratio | 95% CI | Unseen |
| Policy-only | 0.918 | [0.881, 0.956] | 0.934 | 1.009 | [0.986, 1.025] | 1.076 |
| Initial proposal | 0.922 | [0.878, 0.967] | 0.939 | 1.058 | [1.020, 1.093] | 1.107 |
| Recursive analysis | 0.929 | [0.897, 0.958] | 0.958 | 0.999 | [0.990, 1.009] | 1.003 |
| Fixed plan | Seed A | Seed B |
|---|---|---|
| Range4 | 1.000 | 1.000 |
| Range8 | 1.030 | 0.967 |
| K2p2 | 1.020 | 1.091 |
| Fused | 1.070 | 1.210 |
| Bank oracle | 0.889 | 0.848 |
| Response (MiB/s) | S / Uniform | S / Balanced |
|---|---|---|
| Local | 0.985 | 0.991 |
| 16 | 0.947 | 1.002 |
| 6 (unseen) | 1.000 | 1.683 |
| 4 | 0.971 | 1.516 |
| 2 | 1.005 | 1.008 |
| 1.5 (unseen) | 0.998 | 1.020 |
| Input | ||
|---|---|---|
| Local | 16/16 | 16/16 |
| 4 MiB/s | 16/16 | 14/16 |
| 1.5 MiB/s | 16/16 | 10/16 |
| Task | Local | 4 MiB/s | 1.5 MiB/s |
|---|---|---|---|
| 01 | 1.051 | 0.843 | 0.609 |
| 02 | 1.036 | 0.805 | 0.599 |
| 03 | 1.005 | 0.718 | Fail |
| 04 | 0.914 | Fail | Fail |
| 05 | 1.035 | 0.645 | 0.447 |
| 06 | 0.991 | 0.761 | Fail |
| Input | Full metric pairs / 16 | SSIM range | PSNR range (dB) |
|---|---|---|---|
| Local | 16/16 | 1.0000 | |
| 4 MiB/s | 10/16 | 0.9843–0.9956 | 32.8–46.4 |
| 1.5 MiB/s | 6/16 | 0.9843–0.9956 | 40.5–46.4 |
| A100 host | RTX3090 host | RTX4090 host | |||
|---|---|---|---|---|---|
| Policy | CPU drift | CPU drift | Stable | CPU drift | Export drift |
| Native concurrency three | 543.25 | 268.55 | 284.64 | 325.80 | 323.66 |
| DAG rule | 370.92 | 236.92 | 235.85 | 293.76 | 282.91 |
| CPUs | Forced plan | Original (s) | Revised (s) |
|---|---|---|---|
| 4 | Parallel | 119.97 | 82.96 |
| 4 | Direct | 94.09 | 68.82 |
| 16 | Direct | 94.54 | 67.33 |
| 16 | Parallel | 106.13 | 71.84 |
| Total worker time | 414.73 | 290.95 | |
| Task | CPUs | Prior (s) | Feedback (s) |
|---|---|---|---|
| Mixed, 6 clips | 4 | 23.22 | 23.48 |
| Mixed, 6 clips | 16 | 9.68 | 10.17 |
| Landscape, 12 clips | 4 | 60.32 | 43.26 |
| Landscape, 12 clips | 16 | 20.70 | 16.45 |
| Policy | Deliveries | Latency / |
|---|---|---|
| No-feedback | 0/3 | Undefined |
| Feedback | 3/3 | 1.020 |
| Reference | 3/3 | 1.000 |
Agent Harness Optimization
Agent harness optimization improves the prompts, tools, memory, workflows, and control code that shape how AI agents act.
See all 204 papers