ARGUS — AI-Agent Reverse Engineering Platform

    1

    Point your AI agent at any website, APK, binary, or firmware image and get an evidence-backed blueprint of how it works — then a clean-room rebuild scaffold to build your own version.

    $29

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    ARGUS — AI-Agent Reverse Engineering Platform

    Example session with this skill installed

    Analyze the APK at ./app-release.apk and give me a security overview and risk score.

    • Read your context and instructions
    • Compiled the argus

    I have analyzed the provided APK file. The analysis covered permissions, embedded SDKs, and hardcoded network strings to determine the overall risk profile.

    APK Analysis Report: app-release.apk

    MetricValue
    Package Namecom.example.analytics.app
    Risk Score75 / 100
    Risk LevelHIGH

    Findings Summary

    • Dangerous Permissions: READ_SMS, ACCESS_FINE_LOCATION, RECORD_AUDIO.
    • Embedded Trackers: Google Ads, Firebase, Facebook SDK, AppsFlyer.
    • Hardcoded URLs: http://dev-api.internal.net/debug, https://telemetry.example.io/v1.

    Detailed Breakdown

    The risk score of 75 is driven primarily by the request for SMS and audio recording permissions which are considered high-sensitivity. The presence of four distinct tracking SDKs adds significant telemetry overhead.

    Next steps

    1. Run argus secrets ./app-release.apk to check for hardcoded API keys in the binary.
    2. Review the READ_SMS requirement with the development team to see if it can be replaced with the SMS Retriever API.
    3. Use argus rebuild ./app-release.apk to generate a scaffold for a privacy-first alternative.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    ARGUS — AI-Agent Reverse Engineering Platform

    Price: $29 one-time purchase

    One-line description

    Point your AI agent at any website, APK, binary, or firmware image and get an evidence-backed blueprint of how it works — then a clean-room rebuild scaffold to build your own version.

    What it is

    ARGUS is a reverse-engineering platform built for AI agents. Instead of a human staring at decompiled code, your agent asks ARGUS and gets answers with evidence attached: what tech stack a site runs, which API endpoints its JavaScript calls, which permissions and tracker SDKs an APK hides, what strings a binary contains, what a firmware image unpacks to, and whether any credentials leaked into the code. Then, in one shot, ARGUS turns that blueprint into a starter codebase — original code you own, not a copy.

    What the buyer gets

    • The complete ARGUS tool (Python 3.9+, MIT-licensed source by Prabhjot Singh) — CLI, MCP server, dashboard, and all engines, ready to install with pip install -e .
    • A ready-to-use SKILL.md — install, CLI reference, MCP tool list, workflows, and honest limitations, written so any AI agent can use ARGUS without guessing
    • 8 MCP tools for any MCP-capable agent (Claude Code, Cursor, Codex, Gemini): web.analyze, binary.info, apk.overview, frida.ps, firmware.scan, rebuild.scaffold, auto.rebuild, secrets.scan
    • Full CLI — scan-web, apk, binary, firmware, ps, secrets, rebuild, auto, setup, dashboard, doctor, tools, serve
    • 6 analysis engines — deep web analyser (pure Python, no browser needed), APK analyser with a 0–100 risk score, radare2 binary analysis, Binwalk firmware scanning, Frida process listing, and a secrets scanner covering 8 credential patterns with masked evidence
    • One-shot rebuild — argus auto <URL> goes from URL to full blueprint + rebuild scaffold in a single command
    • One-command agent setup — argus setup wires ARGUS into Claude Code, Cursor, Codex, and Gemini CLI at once (--dry-run previews the changes first)
    • Local web dashboard — a dark investigation board at http://localhost:8765, zero extra dependencies
    • Agent workflows — trace-feature, map-api-surface, and patch-diff playbooks, plus a worked example agent session
    • Docs — architecture guide, per-agent setup guide, Termux (Android) guide
    • Offline unit tests — python -m pytest tests/ -q
    • Termux-ready — the core runs on Android with just Python; radare2 and Binwalk install via pkg

    Who it is for

    AI-agent builders, security researchers, app analysts, and developers who want to understand how software works — and rebuild it cleanly as their own.

    Honest limitations

    Static web analysis only (no browser/JS execution); the rebuild scaffold is a starting point, not a 1:1 clone; binary analysis needs radare2, firmware needs Binwalk, dynamic analysis needs Frida (attach-mode tracing not in this version); no iOS IPA support and no Ghidra engine yet. The secrets scanner is pattern-based and can miss non-standard formats. Full details in SKILL.md.

    Licence & legal

    MIT licence. Built for analysis and clean-room reimplementation only — no cracks, keygens, licence bypasses, or DRM circumvention. Respect licences and terms of service. Original work by Prabhjot Singh — https://github.com/Thegreatprabh/Argus

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Agents & Orchestration.

    Static analysis only; no JS execution or authenticated flows. Binary/firmware engines require radare2/binwalk on PATH. No iOS IPA or Ghidra support. Pattern-based secret scanning can have false hits.

    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 buying a skill.

    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

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed today

    Software
    PythonGit

    Permissions

    Terminal / Shell
    Write Files
    Network Access
    Read Files
    Listed1 day ago

    What's inside

    Frequently Asked Questions

    This skill is designed for developers, security researchers, and engineers who need to analyze legacy apps, audit third-party APKs for security risks, or map undocumented API endpoints without manual guesswork.

    The skill is compatible with Claude Code, Cursor, and the Gemini CLI, and it integrates seamlessly with any agent supporting MCP tools for automated analysis workflows.

    You will receive the core Python-based analysis engine, the CLI interface, MCP tool definitions, and a documentation guide to help you integrate it with your preferred AI agent.

    Yes, for binary and firmware analysis, you must have Radare2 and Binwalk installed on your system PATH; however, standard web and APK scanning works out-of-the-box via Python.

    It uses static analysis to extract data directly from source files, JS bundles, and binaries, providing evidence-backed facts that prevent the hallucinations common when asking an LLM to guess an architecture.

    The skill currently focuses on static analysis and does not execute JavaScript or handle authenticated web flows, meaning it cannot bypass login screens or capture dynamic data generated during runtime.

    New in Agents & Orchestration