Polish only

ekorbia-desktop

github-repo

A native desktop chat and productivity environment for local AI models

by Erik 2y account · 2 repos·★ 1 · MIT · pushed 2d ago· github.com ·audited by @Erik
14/17
checks passing

Launch-ready. 3 polish items left if you want them.

Sign in to re-audit Re-audits need a free account. All audits of this project →

Launch Certified

No blockers, no fix-before-launch issues. Embed this badge on your README or site — it links back here and stays honest: it always reflects your latest audit.

Launch Certified by FuzzPage
Markdown [![Launch Certified by FuzzPage](https://fuzzpage.com/badge/ekorbia-ekorbia-desktop.svg)](https://fuzzpage.com/p/ekorbia-ekorbia-desktop)
HTML <a href="https://fuzzpage.com/p/ekorbia-ekorbia-desktop"><img src="https://fuzzpage.com/badge/ekorbia-ekorbia-desktop.svg" alt="Launch Certified by FuzzPage"></a>

Polish 3

#16 Search-engine basics Polish

2 SEO gaps: title over 70 chars (search results truncate it); meta description over 160 chars (truncated in results)

Do this

Fix: shorten the <title> to ≤60 chars; trim the meta description to ≤160 chars.

Why we flagged this

Search engines and link-preview crawlers rely on these head tags and crawl directives to index your page and render its search snippet. Each gap quietly lowers how often you're found and how you look when you are.

#17 AI & agent discoverability Polish

Partial AI discoverability — has structured data (JSON-LD), missing llms.txt

Do this

Add a /llms.txt describing the project (llmstxt.org).

Why we flagged this

AI answer engines (ChatGPT, Perplexity, Google AI Overviews) and coding agents increasingly summarize and recommend tools to people who never visit your page. With no machine-readable description, they fall back to guessing from raw HTML — and are likelier to describe you vaguely or wrongly. Your robots.txt has no explicit AI-crawler policy — optional, but worth a deliberate decision.

#18 Alternatives & differentiation Polish LLM

Visitors will compare you to ChatGPT / OpenAI, Ollama, LM Studio, Claude (Anthropic) — your copy doesn't say how you differ

Do this

Add a short "how we're different" note (a sentence, or a small comparison) addressing: ChatGPT / OpenAI, Ollama, LM Studio, Claude (Anthropic).

Why we flagged this

Alternatives a visitor would weigh, and an angle for each: - ChatGPT / OpenAI — Entirely local, offline-first execution with no cloud dependency, API keys, or telemetry—trades cloud scalability for privacy and control - Ollama — Higher-level native desktop UI and integrated productivity features (indexing, multi-tab, prompt management) layered on top of Ollama's command-line model runner - LM Studio — Native desktop chat focused on simplicity; Ekorbia differentiates via deeper productivity integration (file indexing, multi-tab workflows) and cross-platform support - Claude (Anthropic) — Cloud-based conversational AI; Ekorbia trades convenience and model capability for guaranteed privacy, zero telemetry, and full local control The landing copy emphasizes 'no cloud, no API keys, no telemetry' but does not explicitly name or position against known alternatives; it asserts benefits rather than comparative positioning.

Passing 14
Skipped — not applicable 2