Paper 2601.16746
SWE-Pruner: Self-Adaptive Context Pruning for Coding Agents
- Published
- Jan 2026
- Research lab
- Independent
- Citations
- 26
- GitHub
- 304 stars
01 In brief
Summary
SWE-Pruner is a self-adaptive context pruning framework for coding agents, addressing the high API costs and latency from long interaction contexts.
It uses a lightweight 0.6B neural skimmer, trained on 61K synthetic samples, to perform task-aware, line-level pruning based on a goal hint provided by the agent.
Evaluated on multi-turn tasks (SWE-Bench Verified, SWE-QA) and single-turn tasks (Long Code Completion, Long Code QA), SWE-Pruner achieves 23-54% token reduction on agent tasks while improving success rates by 1.2-1.4 percentage points, and up to 14.84x compression on single-turn tasks with minimal performance impact.
It outperforms baselines like LLMLingua2 and RAG by preserving syntactic structure and retaining task-relevant details.
The framework integrates as middleware, requiring minimal modifications to existing agent workflows.
02 From the paper
Abstract
LLM agents have demonstrated remarkable capabilities in software development, but their performance is hampered by long interaction contexts, which incur high API costs and latency. While various context compression approaches such as LongLLMLingua have emerged to tackle this challenge, they typically rely on fixed metrics such as PPL, ignoring the task-specific nature of code understanding. As a result, they frequently disrupt syntactic and logical structure and fail to retain critical implementation details. In this paper, we propose SWE-Pruner, a self-adaptive context pruning framework tailored for coding agents. Drawing inspiration from how human programmers "selectively skim" source code during development and debugging, SWE-Pruner performs task-aware adaptive pruning for long contexts. Given the current task, the agent formulates an explicit goal (e.g., "focus on error handling") as a hint to guide the pruning targets. A lightweight neural skimmer (0.6B parameters) is trained to dynamically select relevant lines from the surrounding context given the goal. Evaluations across four benchmarks and multiple models validate SWE-Pruner's effectiveness in various scenarios, achieving 23-54% token reduction on agent tasks like SWE-Bench Verified while even improving success rates, and up to 14.84x compression on single-turn tasks like LongCodeQA with minimal performance impact.