claude-code-gui

github.com/markes76/claude-code-gui

2026-01-11 ~ 2026-03-12 · 60 days

Ambition Burnout

AI-driven scope creep with human inability to ship

An AI built a palace, a human couldn't open the doors

Death Type

AI-Augmented Scope Paralysis

This project, heavily driven by 'Claude' (78% of commits), expanded its scope with a 'linkedin-cli' sub-project, adding a +570-line README before the initial GUI wrapper was complete. The ambition, amplified by AI, led to 46 commits and 6 unmerged branches, but zero official releases, proving that an AI can build a feature-rich ghost ship.


Cause of Death

1. AI as the primary author

Of the 46 total commits, 'Claude' was explicitly listed as the author for 36 commits, representing approximately 78% of the project's development. The human contribution was primarily supervisory.

2. Rapid, uncontained scope expansion

Initially focused on Claude's CLI, the project rapidly incorporated a new 'linkedin-cli' sub-project, evidenced by a +570-line README.md, before the core functionality was stabilized or released.

3. Zero official releases despite substantial work

Despite accumulating 46 commits over 60 days, and significant development like +513 lines in StreamPage.tsx, the project never achieved an official release. Furthermore, 6 unmerged branches suggest substantial unfinished features.


Vibe Score

54/ 100

Real author was an AI


What They Did

This project commenced as a 'comprehensive desktop GUI wrapper' for the Claude Code CLI, aiming to visualize 'various aspects' of AI interaction. It quickly expanded its ambition to include a 'linkedin-cli' sub-project, demonstrating a TypeScript-first approach with 16 distinct @radix-ui packages.

TypeScript@radix-uitailwindcssnode-ptyxterm

Burnout Analysis

The human developer's burnout score was a remarkably low 20/100, largely due to 'Claude' authoring 78% of the 46 commits. The project's average of 0.72 commits per day over 60 days indicates consistent AI output, not human struggle. The human's true effort was in guiding the AI, not burning the midnight oil.


Dependency Archaeology

A desktop GUI wrapper for an AI CLI, designed for terminal emulation with `node-pty` and `xterm`, amassed 43 dependencies, including 16 distinct `@radix-ui` packages. This impressive, modern stack was meticulously assembled over 60 days, yet never delivered a single official release, demonstrating that a robust build environment is no guarantee of user acquisition.


Autopsy: File Structure

├──src/renderer/src/pages/StreamPage.tsx+513 lines added, -338 lines removed — active development on a streaming interface that streamed to no one.
├──linkedin-cli/README.md+570 lines added, 0 lines removed — the ambitious blueprint for an entirely new sub-project, fully documented but never deployed.
├──src/main/stream-bridge.ts+442 lines added, -68 lines removed — core stream handling logic meticulously crafted for an audience that never materialized.
├──src/renderer/src/pages/AnalyticsPage.tsx+486 lines added, -21 lines removed — an analytics dashboard built to track usage that remained at zero.
├──linkedin-cli/src/api/types.ts+358 lines added — extensive type definitions for an API that shipped to zero users, ensuring type safety for the void.
├──package.json43 dependencies, including 16 @radix-ui packages — a robust ecosystem for a project with no official releases.
└──tsconfig.jsonTypeScript's strictness ensured the code was perfectly type-safe, if only for the developer's local machine.

Eulogy Stats

Total Commits
46
Ambitious Adjectives
4
Deploy Config
No
Estimated Users
0 (despite 46 commits)

Last Words

feat: add linkedin-cli

May your next AI-assisted endeavor prioritize shipping over ever-expanding features.

Ambition BurnoutAI-driven scope creep with human inability to ship

Share on X