Good recovery answers include “not yet.”

No. KodeBack is intended for software the operator owns or is explicitly authorized to inspect and recover. It is positioned around ownership recovery, vendor handoff, modernization, technical diligence, and controlled rebuild—not unauthorized duplication.

Not necessarily, and that is not the useful standard. Minification and compilation may permanently remove original local names, comments, file boundaries, types, and unused code. KodeBack aims to recover a maintainable candidate that preserves declared observed behavior and makes uncertainty explicit.

It can be clearer or better organized, but cleanup and architecture are separate from historical recovery. KodeBack should not invent a preferred architecture merely to claim the source is clean.

Good source maps can recover authored file names, source text, and mappings. They do not automatically solve dependency provenance, private packages, build configuration, assets, runtime authority, external services, or behavior validation.

KodeBack can use readable-bundle lifting, unpacking, static evidence, runtime observation, dependency fingerprints, scene/visual grounding, and bounded semantic judgment. Recoverability and cost vary; some regions may remain retained or unresolved.

Not when the backend implementation was never shipped or otherwise lawfully provided. Client artifacts may reveal endpoints, request shapes, response shapes, identifiers, and assumptions, but not private server implementation.

Not always. The assessment may show a rewrite is preferable. KodeBack is valuable because it can preserve behavior and evidence before that decision and recover reusable source where supported.

The current research architecture covers web applications, Electron applications, and browser extensions, with deeper support for JavaScript/TypeScript-family client artifacts and interactive/WebGL applications. Actual availability depends on the application profile and product stage.

It means comparing configured donor and candidate observations for declared states. Depending on the application, that may include console, network, DOM, accessibility, geometry, pixels, performance, scene state, WebGL, Electron host, extension context, or custom assertions.

KodeBack is research-stage. Kalu Kode is opening a limited number of paid assessment conversations while continuing internal development and owned demonstrations.