GodEyes

    1

    Turn repository code into interactive flow diagrams with source evidence, confidence labels, and review findings. Understand JavaScript/TypeScript applications and PHP backends through a localhost canvas.

    Free

    Security scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    About this skill

    See how your code works

    GodEyes is a free, open-source agent skill that turns repository code into interactive flow diagrams with source evidence, confidence labels, and review findings. It supports JavaScript/TypeScript applications, including PHP backends, and displays reports in a localhost viewer built with Bun, Next.js, React Flow, and shadcn.

    What you get

    • Trace a feature such as a login flow across your repository.
    • Inspect steps and connections with source excerpts and line numbers.
    • Explore Canvas, Findings, and Overview views.
    • Add notes, arrange the diagram, and export the report with annotations.
    • Review snapshot freshness, file hashes, source ranges, and graph validation.

    Install and use

    Requires Python 3.9+ and Bun. Download and extract the repository, then run python3 scripts/install.py. The installer copies the skill and viewer template into ~/.agents/skills/godeyes; use --destination to choose another location.

    In Codex, open the repository you want to review and ask: Use $godeyes to trace and review this repository's login flow. Replace login flow with the feature you want to understand.

    GodEyes reads the source, creates an evidence-backed report, and opens a localhost canvas. The analyzed application is never executed or modified. Reports are snapshots; invoke the skill again after changing the source. Confidence labels describe source support, not runtime verification.

    Released under the MIT License. Source and documentation: https://github.com/ShaungBhone/godeyes

    Static source analysis only; the analyzed application is never executed or modified. Reports are snapshots and must be regenerated after source changes. Confidence indicates source support, not runtime verification. External services remain explicit analysis boundaries. Documented scope: JavaScript/TypeScript applications, including PHP backends.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 15 days ago

    • Free with an Agensi account

    Requires

    Python 3.9+
    Bun
    Codex
    Local web browser
    Python
    Git

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Browser
    Network Access

    Reads the repository under review and writes generated reports, annotations, and the local viewer installation. Shell access runs bundled Python/Bun helpers. Network access is needed to obtain dependencies; the viewer serves locally on 127.0.0.1. The analyzed application is never executed or modified. No connected third-party accounts are documented as required.

    Listed15 days ago

    What's inside

    Frequently Asked Questions

    GodEyes includes source excerpts, line numbers, confidence labels, and snapshot validation for file hashes, source ranges, and graph structure. Confidence describes source support rather than runtime verification; review findings against the cited code.

    The README documents support for JavaScript/TypeScript applications, including PHP backends.

    No. GodEyes reads source code without executing or modifying the analyzed application. Python 3.9+ and Bun are required for the bundled tooling and localhost viewer.

    The MIT-licensed repository includes reusable agent skill instructions, Python snapshot and validation helpers, a Next.js localhost viewer, and the report contract.

    External systems remain explicit analysis boundaries. GodEyes can document source-supported calls and connections, but it does not verify the runtime behavior of external services.

    The README documents installation and invocation with Codex. Install Python 3.9+ and Bun, run python3 scripts/install.py, then invoke $godeyes from a Codex chat opened on the repository to review.

    New in Code Quality & Review

    Free