Case study
REA in Claude Code: an isolated static-analysis A/B trial
A Japanese field report connects REA 6.4.0 to Claude Code in a separate project directory, denies 43 of its 170 tools and common shell launch forms, and compares REA-enabled and REA-disabled runs on two real tasks: an Electron terminal’s packaged files and a saved HAR capture. Both arms answered correctly, while the REA arm took up to 2.00× the time and 1.67× the API-equivalent cost. It is an analysis-evaluation and safety-configuration case, not a reconstruction.
- Targets
- App · Website
- Category
- Community REA investigations
- Relationship
- Author-reported mixed workflow
- Sources reviewed
Requirements & limits
- Scope: The article by keigoly was published on Zenn on 2026-10-14 (JST). It pins REA 6.4.0 from the rea-agents package and runs it with a checksum-verified official Node.js 24.21.0 placed inside the project, because the host’s Homebrew Node 25.8.1 is outside REA’s supported range. All trials used non-interactive claude -p with Claude Opus 5.5 and permission prompts bypassed; the Claude Code version is not stated.
- Limits: Isolation is configuration, not a sandbox: REA is enabled only through a project-local .mcp.json, a generated deny list blocks 43 tools selected from effect metadata, and Bash prefix rules block common package-runner launches. The author states that deny rules are a fence rather than isolation, that other invocation forms bypass the prefix rules, and that prompt injection through analysis output was not tested.
- Scope: The tasks were MIT-licensed Orca terminal v1.4.224 (128 MB of distributed Electron files) and a 2.5 MB HAR from an unnamed venue website; answers were checked against Orca’s public source and the HAR data. The author excluded DRM or license circumvention, online multiplayer games, business or customer data and malware from scope.
- Limits: Each condition ran once (n=1), the REA arm was told to prefer REA tools, parallel processes make timings approximate, and only static JavaScript/web work with one model was tested; an interrupted run was excluded and repeated. One REA analysis result exceeded the 10 MB MCP response limit. Scripts and JSONL logs are excerpted rather than published.
- Attribution: The author reports that REA helped with extraction, such as asar unpacking and HAR element extraction, while grep and small scripts carried code reading in both arms. Nothing was reconstructed: the HAR task identified an existing request path. Native binaries and APKs are deferred to a planned second step; Orca, the venue site, REA and the model service keep their own terms.
About this entry
The description was reviewed against primary sources.
Primary sources
More in Websites
- Dinosaur runner: an REA-inspected speed rule rebuilt with Codex · Case study
- Terminal: a Rust terminal informed by TradingView chart research · Case study
- BaoCut Design: porting reconstructed canvas and design-system editors · Case study
- whatsapp-web-reveng: documenting the legacy WhatsApp Web protocol · Case study
- Ruffle: a Rust reimplementation of Adobe Flash Player · Case study
- Littlebird.ai static home-page rebuild · Case study