BootstrapAgent: Distilling Repository Setup into Reusable Agent Knowledge

2026-05-15

AI Summary

This work addresses the high cost and poor reusability of environment bootstrapping when code agents encounter new repositories. To overcome this, the authors propose BootstrapAgent, which reframes the bootstrapping process as a reusable knowledge problem and introduces a multi-agent collaboration framework to generate structured .bootstrap contracts that enable persistent, verifiable, and consumable bootstrapping knowledge. The approach innovatively incorporates warm repair with clean replay and delta repair with sanity checks to ensure reproducible cold starts while preventing reward hacking. Experimental results demonstrate that BootstrapAgent achieves a 92.9% success rate across three benchmarks—surpassing baseline methods by over 10%—and reduces downstream agent token consumption by 25.9% and build time by 22.3%.

Abstract

Code agents increasingly help developers work with unfamiliar repositories, but every such task depends on a costly prerequisite: bootstrapping the repository into a usable development state. This process requires substantial trial-and-error exploration, yet the resulting knowledge--resolved dependencies, repair strategies--stays trapped in a single conversation, unavailable to future agents. We therefore formulate repository bootstrapping as a reusable startup knowledge problem and introduce BootstrapAgent, a multi-agent framework that distills the heuristics discovered during bootstrap exploration into a persistent, verifiable, agent-consumable .bootstrap contract. Through evidence extraction, structured planning, deterministic Docker-based verification, and trace-driven repair, BootstrapAgent generates a contract covering environment setup, diagnostic checks, minimal verification, and accumulated repair knowledge. We further propose warm repair with clean replay to accelerate iterative debugging without sacrificing cold-start reproducibility, and a delta repair with sanity check to prevent reward hacking. Experiments on three benchmarks show that BootstrapAgent achieves a 92.9% success rate, outperforming the baseline by over 10% while reducing downstream agent token usage by 25.9% and build time by 22.3%. Our code is available at https://github.com/Vossera/BootstrapAgent.

Problem

• repository bootstrapping
• reusable knowledge
• code agents
• development environment setup
• dependency resolution

Innovation

• repository bootstrapping
• multi-agent framework
• reusable agent knowledge
• deterministic verification
• trace-driven repair