# AI tooling for process methods

These methods run on human judgment, so AI is taking a turn at the
judging rather than automating a check. The field moves quickly, and
these entries may be out of date.

## Code review

The built-in reviewers (Claude, Copilot, Cursor) and the
open-source [PR-Agent](https://github.com/The-PR-Agent/pr-agent) (from Qodo, whose proprietary
Qodo Merge continues it) comment at PR time. The empirical
finding (Rasheed et al. 2024)[^rasheed2024]: LLM reviewers catch a meaningful subset
of defects but miss design-level issues, consistent with
[Code review](https://quality.stereobooster.com/code-review.md).

## Exploratory testing

General-purpose vision-LLM browser agents (Claude / OpenAI
computer use, the open-source [Browser Use](https://github.com/browser-use/browser-use))
extend happy-path coverage. Research benchmarks: **WebVoyager**
(He et al. 2024)[^he2024], **AppAgent** (Zhang et al. 2023)[^zhang2023], **VisualWebArena**
(Koh et al. 2024)[^koh2024]. These agents substitute poorly for skilled
exploratory testing.

## Refactoring practice

General-purpose agents (Cursor, Claude Code, Aider) are the
workhorse; JetBrains AI Assistant augments IntelliJ's mechanical
refactorings. Structured bulk refactoring uses a deterministic
rule engine as the oracle: [OpenRewrite](https://docs.openrewrite.org/). Research:
[SWE-agent](https://swe-agent.com), [AutoCodeRover](https://autocoderover.dev) for
repo-level autonomous changes. CodeScene's hotspot core is
legitimate; its smell-catalog layer is folklore-tier per
[Git-history hotspots](https://quality.stereobooster.com/git-hotspots.md).

## Requirements quality checking

Requirements tooling splits by activity: *gathering* requirements
from informal input has usable tools, while auditing existing
requirements for ambiguity and contradiction is mostly research.
The tools and the published results are in
[AI tooling for requirements engineering](https://quality.stereobooster.com/ai-requirements-tools.md).

## References

[^rasheed2024]: Rasheed, Zeeshan, Malik Abdul Sami, Muhammad Waseem, et al. 2024. *[AI-powered Code Review with LLMs: Early Results](https://arxiv.org/pdf/2404.18496)*. <https://doi.org/10.48550/arXiv.2404.18496>.
[^he2024]: He, Hongliang, Wenlin Yao, Kaixin Ma, et al. 2024. *[WebVoyager: Building an End-to-End Web Agent with Large Multimodal Models](https://arxiv.org/pdf/2401.13919)*. <https://doi.org/10.48550/arXiv.2401.13919>.
[^zhang2023]: Zhang, Chi, Zhao Yang, Jiaxuan Liu, et al. 2023. *[AppAgent: Multimodal Agents as Smartphone Users](https://arxiv.org/pdf/2312.13771)*. <https://doi.org/10.48550/arXiv.2312.13771>.
[^koh2024]: Koh, Jing Yu, Robert Lo, Lawrence Jang, et al. 2024. "[VisualWebArena: Evaluating Multimodal Agents on Realistic Visual Web Tasks](https://arxiv.org/pdf/2401.13649)." *Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)*, 881–905. <https://doi.org/10.18653/v1/2024.acl-long.50>.