GitHub Integrity Guard 作者: Ahmed Samy Elkhouly
Provides instant transparency regarding the authenticity and health of GitHub repositories.
擴充套件後設資料
畫面擷圖
關於此擴充套件
GitHub Integrity Guard is an advanced, open-source browser extension architected to eliminate star-farming deception. It injects a real-time, highly accessible scorecard directly into GitHub repositories, analyzing authentic community health and protecting independent developers from bot-driven metrics.
Going entirely against the current trend, this tool is 100% AI-free. There are no black-box LLMs, no hallucinations, and no data tracking. It relies entirely on deterministic math applied directly to GitHub's raw API to provide instant, objective transparency.
How It Works (The Engine & UI)
Token-Powered API: Securely utilizes your GitHub Personal Access Token (PAT) via local AES-GCM encryption to bypass strict rate limits.
Smart UI Injection: Seamlessly injects a clean Scorecard into the native GitHub sidebar. If the sidebar is missing, a resilient Floating Card appears.
Visual Transparency: Displays a definitive Trust Score (out of 100) using a traffic-light color system: Red (Danger), Yellow (Warning), and Green (Healthy).
Full Breakdown: Exposes the raw numbers underneath the score: Fork-to-Star ratio, 90-day Commits, Issue Health, and Discussion Depth.
24-Hour Smart Cache: Successful audits are cached locally for 24 hours to preserve your API limits.
The Tiered Scoring Architecture
To prevent penalizing new or niche projects, the algorithm dynamically adjusts its rules based on the repository's total stars:
Tier 1 (Incubator | < 200 Stars): Baseline 90+ score. Penalized only if absolutely dormant for > 365 days.
Tier 2 (Growth | 200 - 5,000 Stars): Pulse-focused auditing. Rewards active commit cadence and community issue discussions over raw fork numbers.
Tier 3 (Established | 5,000 - 10,000 Stars): Balanced auditing. Matches organic community adoption with the project's growing popularity.
Tier 4 (High-Traffic | > 10,000 Stars): Strict anti-fraud enforcement. Hard-caps scores at 30 (Danger Zone) if the fork ratio falls below 3% to kill bot farms.
Uncompromised Accessibility (WCAG 2.1 Compliant)
Engineered natively for screen-reader users, this tool actively rejects cluttered charts in favor of pure semantic data:
Instant Audio Briefing: A visually-hidden aria-live region reads a concise summary of the Trust Score directly upon load.
Accessible Export: A dedicated native button allows you to copy the entire metrics breakdown to your clipboard with an instant audio confirmation.
Keyboard Native: Every element is strictly navigable via standard keyboard inputs.
Initial Setup & Privacy (Zero Telemetry)
Step 1: Generate a minimal-scope GitHub Personal Access Token (or run gh auth token).
Step 2: Open the extension's Options page, paste your token, and save.
Military-Grade Local Encryption: Your token is immediately encrypted locally using AES-GCM via the native Web Crypto API.
Zero Telemetry: Your data never leaves your device. We do not track you, and there are zero external servers involved.
Going entirely against the current trend, this tool is 100% AI-free. There are no black-box LLMs, no hallucinations, and no data tracking. It relies entirely on deterministic math applied directly to GitHub's raw API to provide instant, objective transparency.
How It Works (The Engine & UI)
Token-Powered API: Securely utilizes your GitHub Personal Access Token (PAT) via local AES-GCM encryption to bypass strict rate limits.
Smart UI Injection: Seamlessly injects a clean Scorecard into the native GitHub sidebar. If the sidebar is missing, a resilient Floating Card appears.
Visual Transparency: Displays a definitive Trust Score (out of 100) using a traffic-light color system: Red (Danger), Yellow (Warning), and Green (Healthy).
Full Breakdown: Exposes the raw numbers underneath the score: Fork-to-Star ratio, 90-day Commits, Issue Health, and Discussion Depth.
24-Hour Smart Cache: Successful audits are cached locally for 24 hours to preserve your API limits.
The Tiered Scoring Architecture
To prevent penalizing new or niche projects, the algorithm dynamically adjusts its rules based on the repository's total stars:
Tier 1 (Incubator | < 200 Stars): Baseline 90+ score. Penalized only if absolutely dormant for > 365 days.
Tier 2 (Growth | 200 - 5,000 Stars): Pulse-focused auditing. Rewards active commit cadence and community issue discussions over raw fork numbers.
Tier 3 (Established | 5,000 - 10,000 Stars): Balanced auditing. Matches organic community adoption with the project's growing popularity.
Tier 4 (High-Traffic | > 10,000 Stars): Strict anti-fraud enforcement. Hard-caps scores at 30 (Danger Zone) if the fork ratio falls below 3% to kill bot farms.
Uncompromised Accessibility (WCAG 2.1 Compliant)
Engineered natively for screen-reader users, this tool actively rejects cluttered charts in favor of pure semantic data:
Instant Audio Briefing: A visually-hidden aria-live region reads a concise summary of the Trust Score directly upon load.
Accessible Export: A dedicated native button allows you to copy the entire metrics breakdown to your clipboard with an instant audio confirmation.
Keyboard Native: Every element is strictly navigable via standard keyboard inputs.
Initial Setup & Privacy (Zero Telemetry)
Step 1: Generate a minimal-scope GitHub Personal Access Token (or run gh auth token).
Step 2: Open the extension's Options page, paste your token, and save.
Military-Grade Local Encryption: Your token is immediately encrypted locally using AES-GCM via the native Web Crypto API.
Zero Telemetry: Your data never leaves your device. We do not track you, and there are zero external servers involved.
由 1 位評論者給出 0 分
權限與資料
必要權限:
- 存取瀏覽器分頁
- 存取您在 github.com 的資料
選用的權限:
- 存取您在 github.com 的資料
- 存取您在 api.github.com 的資料
收集下列資料:
- 開發者聲稱此擴充套件不要求收集任何資料。
更多資訊
- 版本
- 1.0.0
- 大小
- 47.97 KB
- 最近更新
- 5 天前 (2026年5月26日)
- 授權條款
- MIT License
- 隱私權保護政策
- 閱讀此附加元件的隱私權保護政策
- 版本紀錄
- 新增至收藏集
As a blind software engineer, the open-source community is my absolute lifeline. Projects like the NVDA screen reader are the fundamental reason I can use a computer and write code every day. I built this extension to give back to this incredible ecosystem and keep it clean from bot manipulation. Going entirely against the current trend, this tool is strictly AI-free—relying solely on pure math, deterministic logic, and zero-chart accessibility to ensure independent creators get the fair exposure they deserve.