FastContext: Training Efficient Repository Explorer for Coding Agents

Papers with Code Trending Papers

Summary

FastContext introduces specialized exploration models that separate repository exploration from code solving in LLM agents, reducing token consumption by up to 60% while improving resolution rates on software engineering benchmarks.

Large Language Model (LLM) coding agents have achieved strong results on software engineering tasks, yet repository exploration remains a major bottleneck: locating relevant code consumes substantial token budget and pollutes the agent's context with irrelevant snippets. In most agents, the same model explores the repository and solves the task, leaving exploratory reads and searches in the solver's history. We present FastContext, a dedicated exploration subagent that separates repository exploration from solving. Invoked on demand, FastContext issues parallel tool calls and returns concise file paths and line ranges as focused context. FastContext is powered by specialized exploration models spanning 4B--30B parameters. We bootstrap them from strong reference-model trajectories and refine them with task-grounded rewards for broad first-turn search, multi-turn evidence gathering, and precise citation generation. Across SWE-bench Multilingual, SWE-bench Pro, and SWE-QA, integrating FastContext into Mini-SWE-Agent improves end-to-end resolution rates up to 5.5\% while reducing coding-agent token consumption up to 60\%, with marginal overhead. These results show that repository exploration can be separated from solving and handled effectively by specialized models. Code and data: https://github.com/microsoft/fastcontext
Original Article
View Cached Full Text

Cached at: 06/16/26, 02:59 AM

Paper page - FastContext: Training Efficient Repository Explorer for Coding Agents

Source: https://huggingface.co/papers/2606.14066

Abstract

FastContext separates repository exploration from code solving in LLM agents using specialized exploration models that reduce token consumption and improve resolution rates.

Large Language Model(LLM)coding agentshave achieved strong results on software engineering tasks, yetrepository explorationremains a major bottleneck: locating relevant code consumes substantial token budget and pollutes the agent’s context with irrelevant snippets. In most agents, the same model explores the repository and solves the task, leaving exploratory reads and searches in the solver’s history. We present FastContext, a dedicatedexploration subagentthat separatesrepository explorationfrom solving. Invoked on demand, FastContext issues paralleltool callsand returns concise file paths and line ranges as focused context. FastContext is powered byspecialized exploration modelsspanning 4B--30B parameters. We bootstrap them from strongreference-model trajectoriesand refine them withtask-grounded rewardsfor broadfirst-turn search,multi-turn evidence gathering, andprecise citation generation. AcrossSWE-bench Multilingual,SWE-bench Pro, andSWE-QA, integrating FastContext intoMini-SWE-Agentimproves end-to-end resolution rates up to 5.5\% while reducing coding-agent token consumption up to 60\%, with marginal overhead. These results show thatrepository explorationcan be separated from solving and handled effectively by specialized models. Code and data: https://github.com/microsoft/fastcontext

View arXiv pageView PDFProject pageGitHub152Add to collection

Get this paper in your agent:

hf papers read 2606\.14066

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2606.14066 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2606.14066 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2606.14066 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

microsoft/FastContext-1.0-4B-SFT

Hugging Face Models Trending

Microsoft released FastContext-1.0, a lightweight repository-exploration subagent for LLM coding agents that reduces main-agent token consumption by up to 60% while improving resolution rates by up to 5.5%.

SWE-Explore: Benchmarking How Coding Agents Explore Repositories

Hugging Face Daily Papers

SWE-Explore introduces a benchmark for evaluating coding agents' repository exploration capabilities, requiring ranked lists of relevant code regions within line budgets. Experiments show agentic exploration outperforms traditional retrieval, and line-level coverage remains a key differentiator.

LLM Agents Can See Code Repositories

Hugging Face Daily Papers

This paper presents the first systematic empirical study of using visual repository representations to enhance LLM-based coding agents, showing that integrating visual graphs as a supplementary modality reduces token consumption by up to 26% while maintaining or improving issue-resolution accuracy.