REA

• Connect reverse-engineering workflows to coding agents through an MCP server or the REA CLI.
• Use agent-assisted analysis across multiple target types, with capabilities and prerequisites that vary by format.
• Set up supported coding-agent integrations with the project's agent setup command, then review and approve proposed changes.

Developers, security researchers, and technical analysts who already work with coding agents and need a repeatable interface for examining software or application artifacts.

REA is best understood as an agent-facing workflow layer for reverse engineering, not a replacement for specialist analysis tools. It can help connect a coding agent to a target, but supported formats, local dependencies, and the quality of conclusions depend on the analysis path. Teams should verify the setup requirements for their specific target before adopting it.

• Combines an MCP workflow and CLI setup around several coding-agent environments.
• The MIT license permits use and modification subject to its license terms.

• Some analysis paths rely on separately installed tools or runtimes, so setup effort depends on the target format.
• Analysis is mediated through a coding agent; model-provider data handling remains a separate consideration from REA's local analysis claim.

The repository is MIT-licensed and does not list a paid REA subscription. Users still need a compatible Node.js/npm setup and coding-agent access; optional analysis engines, model providers, or other third-party services may have their own terms or costs.

Check the repository license ↗