Squeez: Task-Conditioned Tool-Output Pruning for Coding Agents
For developers of coding agents, this work provides an efficient pruning technique to reduce token consumption, but it is incremental as it fine-tunes an existing small model on a new benchmark.
The paper introduces a task-conditioned tool-output pruning method for coding agents, achieving 0.86 recall and 0.80 F1 while removing 92% of input tokens, outperforming larger zero-shot models by 11 recall points.
Coding agents repeatedly consume long tool observations even though only a small fraction of each observation matters for the next step. We study task-conditioned tool-output pruning: given a focused query and one tool output, return the smallest verbatim evidence block the agent should inspect next. We introduce a benchmark of 11,477 examples built from SWE-bench repository interactions and synthetic multi-ecosystem tool outputs, with a manually curated 618-example test set. We fine-tune Qwen 3.5 2B with LoRA and compare it against larger zero-shot models and heuristic pruning baselines. Our model reaches 0.86 recall and 0.80 F1 while removing 92% of input tokens, outperforming zero-shot Qwen 3.5 35B A3B by 11 recall points and all heuristic baselines by a wide margin.