AgenticTyper: Automated Typing of Legacy Software Projects Using Agentic AI

Fuente: arXiv
Saved in:
Bibliographic Details
Main Author: Pohle, Clemens
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915815298695168
author Pohle, Clemens
author_facet Pohle, Clemens
contents Legacy JavaScript systems lack type safety, making maintenance risky. While TypeScript can help, manually adding types is expensive. Previous automated typing research focuses on type inference but rarely addresses type checking setup, definition generation, bug identification, or behavioral correctness at repository scale. We present AgenticTyper, a Large Language Model (LLM)-based agentic system that addresses these gaps through iterative error correction and behavior preservation via transpilation comparison. Evaluation on two proprietary repositories (81K LOC) shows that AgenticTyper resolves all 633 initial type errors in 20 minutes, reducing manual effort from one working day.
format Preprint
id arxiv_https___arxiv_org_abs_2602_21251
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle AgenticTyper: Automated Typing of Legacy Software Projects Using Agentic AI
Pohle, Clemens
Software Engineering
Artificial Intelligence
Multiagent Systems
Programming Languages
D.2.3; D.2.7; F.3.3; I.2.2; I.2.11
Legacy JavaScript systems lack type safety, making maintenance risky. While TypeScript can help, manually adding types is expensive. Previous automated typing research focuses on type inference but rarely addresses type checking setup, definition generation, bug identification, or behavioral correctness at repository scale. We present AgenticTyper, a Large Language Model (LLM)-based agentic system that addresses these gaps through iterative error correction and behavior preservation via transpilation comparison. Evaluation on two proprietary repositories (81K LOC) shows that AgenticTyper resolves all 633 initial type errors in 20 minutes, reducing manual effort from one working day.
title AgenticTyper: Automated Typing of Legacy Software Projects Using Agentic AI
topic Software Engineering
Artificial Intelligence
Multiagent Systems
Programming Languages
D.2.3; D.2.7; F.3.3; I.2.2; I.2.11
url https://arxiv.org/abs/2602.21251