Skip to content
Back to search

open-reverselab

Open-source AI reverse-engineering agent platform and MCP server for Ghidra, Frida, x64dbg and Rizin — automated PE/APK/binary analysis, CTF and malware research, with 100+ MCP tools and a 194-article runnable knowledge base.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today

Check it deeper on your machine

It has changed since its scan on 11 Nov 2026.

As a Claude Code plugin, install it and then ask Claude Code:

/plugin marketplace add keithmackay/mackayi/plugin install scanskillsafety@mackayiDo a deep scan of https://github.com/ling71671/open-reverselab with scanskillsafety

Or from a terminal, clone it and run it:

git clone https://github.com/keithmackay/scanskillsafety-skill ~/.claude/skills/scanskillsafetynode ~/.claude/skills/scanskillsafety/cli.cjs --deep 'https://github.com/ling71671/open-reverselab'

It checks the exact commit: every file, its packages against OSV (api.osv.dev), how it compares with this page's scan (findsafeskills.com receives the repository's address), and, if you agree, Claude reading every file with your own Claude subscription. It never runs the code.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/ling71671--open-reverselab/badge.svg)](/listings/ling71671--open-reverselab)

HTML

<a href="/listings/ling71671--open-reverselab"><img src="/api/listings/ling71671--open-reverselab/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

Pick the path that matches your role, not your platform — the platform differences are below the role split.

For humans

**Windows (recommended for first run)** — double-click `START_HERE.bat` (or `START_HERE.cmd`) in the repo root. It checks Python, `uv`, Git, the `reverse_lab_tools` MCP server, runs real MCP smoke calls, and writes `reports/misc/first-run-report.json` plus `reports/misc/mcp-smoke-report.json`.

**macOS / Linux** — run `./START_HERE.sh` from the repo root. It does the same first-run checks using POSIX shell wrappers under `tools/bin/`. Windows-only GUI/PE tools are skipped or reported as such.

<details> <summary><b>Per-board install (after first-run passes)</b></summary>

# Pick the boards you actually need — the lab is modular.
.\scripts\misc\bootstrap.ps1                # core script wrappers (no downloads)
.\scripts\misc\install_tools.ps1 -CTF       # Web tools (sqlmap, nuclei, ffuf, jwt_tool, …)
.\scripts\misc\install_tools.ps1 -Android   # APK tools (apktool, jadx, frida, uber-apk-signer, …)
.\scripts\misc\install_tools.ps1 -Windows   # PE tools (cutter, pe-bear, procmon, …)
.\scripts\misc\install_tools.ps1 -Common    # Ghidra + Maven

macOS / Linux equivalent:

./scripts/misc/bootstrap.sh
export PATH="$PWD/tools/bin:$PWD/tools/ctf-website/bin:$PATH"
python scripts/misc/ai_toolcheck.py --board misc    # verify the fresh-clone core

Install only the boards you need. Don't dump the full toolchain if you're only doing web CTF.

</details>

<details> <summary><b>Windows Defender / antivirus note</b></summary>

After installing CTF / ExploitDB tools, Windows Defender may flag exploit samples or payload documentation — e.g. `tools/ctf-website/exploitdb`, `kb/ctf-website/techniques/24-database/03-nosql-injection.md`, `docs/llms-full.txt`. These contain security test payloads, webshells, shellcode, or ExploitDB samples and are **expected components**.

We recommend **minimal path exclusions** rather than excluding the entire repo:

Add-MpPreference -ExclusionPath "D:\open-reverselab\tools\ctf-website\exploitdb"

If individual documents are flagged, handle them case-by-case.

</details>

For AI Agents

1. Clone into a dedicated local directory, e.g. `<workspace>/open-reverselab`. 2. **Claude Code**: `cd <workspace>/open-reverselab`, then start your session. 3. **Codex APP**: Open the existing `open-reverselab` folder directly (no re-clone needed). 4. Automated AI install: send the contents of [`templates/prompts/ai-install.en.md`](templates/prompts/ai-install.en.md) to your Agent. 5. Create a task: `python scripts/misc/new_task.py --board ctf-website --name <name>`. 6. Verify real MCP invocation whenever switching machines or updating configs:

   uv run --project tools/skills/mcp/ReverseLabToolsMCP \
     python scripts/misc/mcp_smoke_check.py --write-report

**Context chain** — Agents load context through this sequence:

CLAUDE.md → AGENTS.md → AI-USAGE.md → boards/<board>/AI-USAGE.md

Use [codex-session-patcher](https://github.com/ryfineZ/codex-session-patcher) for one-click configuration of project-level `.codex/` and MCP servers.

<details> <summary><b>Environment snapshot protocol (host-level, shared across projects)</b></summary>

When first opening this project with an AI Agent, it inspects your local environment (OS, toolchains, reverse-engineering libraries, devices, sanitized env vars, network, workspace) following the protocol in [AGENTS.md](AGENTS.md), and caches a snapshot at `~/.open-reverselab/env/env.md` (`%USERPROFILE%\.open-reverselab\env\env.md` on Windows).

This file is **host-level** and **shared across sessions**: subsequent sessions read it directly without re-probing, unless 7 days have passed or the protocol version has bumped. It remains local and is never committed. Sensitive variables are redacted (keys marked "SET", proxies stripped of credentials).

</details>

Post-install verification

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 162 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. <div align="center">
  2. # Open ReverseLab
  3. <img src="assets/social-preview-v2.png" width="960" alt="Open ReverseLab — AI reverse engineering workspace">
  4. **An AI workspace for target-based reverse engineering**
  5. Desktop workspaces · red / blue / both · Parallel Goals · Knowledge base · MCP
  6. <p><strong>English</strong> · <a href="README.zh.md">简体中文</a></p>
  7. <p><a href="https://github.com/LING71671/open-reverselab/releases/tag/v2.0.0-beta2"><strong>Download for Windows</strong></a> &nbsp; · &nbsp; <a href="desktop/README.md">Desktop guide</a> &nbsp; · &nbsp; <a href="https://ling71671.github.io/open-reverselab/">Website & docs</a> &nbsp; · &nbsp; <a href="https://github.com/LING71671/open-reverselab-knowledge">Knowledge base</a></p>
  8. [![Release](https://img.shields.io/github/v/release/LING71671/open-reverselab?include_prereleases&style=flat-square&color=242424)](https://github.com/LING71671/open-reverselab/releases)
  9. [![License: Apache-2.0](https://img.shields.io/badge/license-Apache--2.0-242424?style=flat-square)](LICENSE)
  10. [![Ask DeepWiki](assets/badges/ask-deepwiki.svg)](https://deepwiki.com/LING71671/open-reverselab)
  11. </div>
  12. ## Sponsors
  13. <table width="100%">
  14. <tr>
  15. <td width="46%" align="center" valign="middle"><a href="https://infistar.cc/register?aff=OPENREVERSELAB&amp;ref_source=link"><img src="assets/sponsors/infistar-banner-en.png" width="420" alt="Infistar.cc — Multi-model API gateway"></a></td>
  16. <td valign="middle"><p><strong><a href="https://infistar.cc/register?aff=OPENREVERSELAB&amp;ref_source=link">Infistar.cc</a></strong></p><p>Claude, GPT, Gemini, Grok and DeepSeek with one key, up to 99% off official prices. 100% genuine models, every model verifiable, with 1,000+ redundant routes and automatic failover. Access 200+ models from 16 brands for AI reverse engineering and MCP workflows.</p><p><a href="https://infistar.cc/register?aff=OPENREVERSELAB&amp;ref_source=link"><strong>Sign up for CNY 5 in free credit</strong></a></p></td>
  17. </tr>
  18. <tr>
  19. <td align="center" valign="middle"><a href="https://sentry.io"><picture><source media="(prefers-color-scheme: dark)" srcset="assets/sponsors/sentry-wordmark-dark.svg"><img src="assets/sponsors/sentry-wordmark-light.svg" width="150" alt="Sentry"></picture></a></td>
  20. <td valign="middle"><p><strong><a href="https://sentry.io">Sentry</a></strong></p><p>Sponsored account for error monitoring and performance tracing across the project toolchain.</p></td>
  21. </tr>
  22. </table>
  23. **Partner pick: [Mozhenzhen](https://mozhenzhen.com/go/open-reverselab).** Check Mozhenzhen before buying AI. Compare 2,000+ models across 350K+ listings, verify models and read user reviews, all for free.
  24. <p align="center"><sub>Sponsor details and ways to support the project: <a href="SPONSORS.md">SPONSORS.md</a></sub></p>
  25. ## Desktop
  26. Select `red / blue / both` in the upper-left corner, choose a workspace, and enter a file, directory, or URL target with a Goal. The live canvas shows execution branches; activity, logs, reports, and health views retain the evidence behind the result.
  27. <img src="assets/desktop/knowledge-light.png" alt="Open ReverseLab Desktop knowledge reader in both mode" width="1280">
  28. - **Independent strategies.** `red` and `blue` use separate analysis policies. `both` lets each strategy plan multiple actions, exchange evidence during execution, and independently check their results before combined verification.
  29. - **Concurrent Goals.** Goals and ready actions run concurrently without an application-level count limit. Switching views, tasks, or workspaces does not redirect a running task's output.
  30. - **Managed workspaces.** First launch creates a default workspace. Open or create others, import samples, and keep original and working copies separate.
  31. - **Goal control.** Goals have no time or token limit by default. Set either limit when needed, stop a run, or add instructions while it is running. Verification results and failed attempts remain in the execution history.
  32. - **Knowledge and usability.** Browse defense articles and paired references, switch light/dark themes, and retain unsent drafts, reading positions, and filters during navigation.
  33. ### Download and setup
  34. [Download v2.0.0-beta2](https://github.com/LING71671/open-reverselab/releases/tag/v2.0.0-beta2) for Windows x64: choose the installer (`-setup.exe`) or portable ZIP. Both packages contain the desktop runtime without a bundled knowledge base or MCP server. Install knowledge through **Extensions** and configure MCP connections separately.
  35. 1. Install Python 3.11 or newer and make the `python` command available on PATH. Windows also needs the Microsoft Edge WebView2 runtime.
  36. 2. Run the installer, or extract the complete portable ZIP and launch `OpenReverseLab.exe`. Keep personal workspaces outside the installation directory.
  37. 3. Open Settings and configure an OpenAI-compatible chat-completions endpoint and model, then enter a target and objective. SHA256 verification can run locally without a model.
  38. Windows x64 is the desktop packaging target. Python and third-party tools such as Ghidra, Frida, and x64dbg are installed separately. macOS/Linux support the source-based CLI/MCP workflow; native desktop packages for those platforms are not currently provided. The installed edition offers Beta and Stable update channels; beta builds select Beta initially. See [application updates](desktop/UPDATES.md).
  39. The desktop executes target triage, Python actions, and tools discovered through configured MCP connections, without a default overall task deadline. Install and configure MCP servers and external tools as needed. Workspace files stay on disk, while model-assisted runs send task context and selected evidence to the endpoint you configure. Python actions run with the current user's permissions. See the [desktop guide](desktop/README.md) for configuration and data handling.
  40. ## Knowledge base and MCP
  41. Knowledge sources live in [open-reverselab-knowledge](https://github.com/LING71671/open-reverselab-knowledge). This repository pins that content through the `kb` Git submodule. For an existing checkout, run `git submodule update --init --recursive`. Submit knowledge changes in the knowledge repository.
  42. The **Extensions** page imports knowledge packs and MCP descriptors from GitHub repositories, catalogs, folders, or ZIP files. Under **Sources**, browse **Open ReverseLab official knowledge** and install the pack; the desktop runtime is distributed separately. Packs can be enabled, disabled, and exported, including selected networks with their referenced documents. MCP connections support stdio and Streamable HTTP. See the [v1 protocol and usage](EXTENSIONS.md) and [authoring skill](tools/skills/create-reverselab-extension/SKILL.md).
  43. The knowledge base covers Web/CTF, Android/APK, Windows/PE, and general topics such as cryptography and protocols. Articles carry `track: red | blue | both`; `both` articles are shared. Related attack and defense articles are linked through index metadata and `defends` references. The current catalog is the source of truth; not every article has a one-to-one counterpart.
  44. | Area | Knowledge entry | Tool examples |
  45. | --- | --- | --- |
  46. | Web / CTF | [Web knowledge](https://github.com/LING71671/open-reverselab-knowledge/blob/main/shared/ctf-website/README.md) | `kb_router`, `http_probe`, `run_ctf_tool` |
  47. | Android / APK | [APK knowledge](https://github.com/LING71671/open-reverselab-knowledge/blob/main/shared/apk-reverse/README.md) | `android_app_baseline`, `android_crypto_unpack_recipe` |
  48. | Windows / PE | [PE knowledge](https://github.com/LING71671/open-reverselab-knowledge/blob/main/shared/pe-reverse/README.md) | `triage_pe`, `ghidra_headless_analyze`, `sample_full_workup` |
  49. | General | [Cross-domain knowledge](https://github.com/LING71671/open-reverselab-knowledge/blob/main/shared/general/README.md) | `solve_crypto_from_evidence`, `rizin_*` |
  50. CLI/MCP track selection defaults to `blue`; select `red` or `both` explicitly through `REVERSELAB_MODE`, `--track`, or the MCP track tools. Desktop selection is scoped to its Goals and does not rewrite another client's environment.
  51. ```sh
  52. python scripts/misc/kb_router.py "JWT" --track blue
  53. python scripts/misc/desktop_kb.py list --track both
  54. ```
  55. ## CLI and AI-agent setup
  56. Clone the repository, then use `START_HERE.bat` / `START_HERE.cmd` on Windows or `./START_HERE.sh` on macOS/Linux. These are setup and MCP checks, not the desktop launcher. The reports are written under `reports/misc/`.
  57. ```sh
  58. git clone --recurse-submodules https://github.com/LING71671/open-reverselab.git
  59. cd open-reverselab
  60. ```
  61. Open this directory in your AI client and read [AI-USAGE.md](AI-USAGE.md). The context chain is `CLAUDE.md → AGENTS.md → AI-USAGE.md → boards/<board>/AI-USAGE.md`. For guided installation, use the [English](templates/prompts/ai-install.en.md) or [Chinese](templates/prompts/ai-install.zh.md) prompt. MCP clients need `uv` and the dependencies in `tools/skills/mcp/ReverseLabToolsMCP`.
  62. ```sh
  63. uv sync --project tools/skills/mcp/ReverseLabToolsMCP
  64. uv run --project tools/skills/mcp/ReverseLabToolsMCP python scripts/misc/mcp_smoke_check.py --write-report
  65. python scripts/misc/lab_healthcheck.py
  66. python scripts/misc/ai_toolcheck.py --board misc
  67. ```
  68. Install only the tool groups needed for the task. Windows provides `scripts/misc/install_tools.ps1 -CTF`, `-Android`, `-Windows`, and `-Common`; POSIX wrappers are generated by `scripts/misc/bootstrap.sh`. See [START.md](START.md) and [tools/README.md](tools/README.md).
  69. ## Build the desktop
  70. With Node 22+, pnpm, Python, Rust, and the platform Tauri prerequisites installed:
  71. ```sh
  72. cd desktop
  73. pnpm install --frozen-lockfile
  74. pnpm tauri dev
  75. # Or create a production executable:
  76. pnpm tauri build --no-bundle
  77. ```
  78. `pnpm dev` is the browser preview. See [development](desktop/DEV_GUIDE.md), [architecture](desktop/ARCHITECTURE.md), [Agent contracts](desktop/AGENT_SYSTEM.md), [error codes](desktop/ERROR_CODES.md), and [release instructions](RELEASING.md).
  79. ## Data and repository layout
  80. | Location | Purpose |
  81. | --- | --- |
  82. | `desktop/` | Desktop UI, queue, execution engine, and Tauri host |
  83. | `kb/` | Git submodule linking the independent knowledge repository |
  84. | `scripts/`, `tools/`, `templates/` | Automation, adapters, and templates |
  85. | `boards/` | Task routing for external AI clients |
  86. | `site/` | Documentation-site source |
  87. | Selected workspace's `.reverselab/` | Desktop originals/work copies, Goal records, exports, and reports |
  88. | Repository `cases/`, `samples/`, `notes/`, `exports/`, `reports/` | Local CLI analysis artifacts; excluded from publication |
  89. Desktop data defaults to `~/.open-reverselab/workspaces/Default`. The host environment snapshot is separate at `~/.open-reverselab/env/env.md`. Neither belongs in a public commit. See [workspace storage](desktop/WORKSPACES.md).
  90. ## Contributing and publication
  91. Read [Contributing](.github/CONTRIBUTING.md), [Security](SECURITY.md), and [Publication](PUBLICATION.md). Publish general implementations and synthetic examples; keep real targets, samples, credentials, and case evidence private. Before submitting, run `python scripts/misc/public_release_check.py` and the checks relevant to the changed module. Release history is in [CHANGELOG.md](CHANGELOG.md).
  92. ---
  93. ## Star History
  94. <div align="center">
  95. <a href="https://star-history.com/#LING71671/open-reverselab&Date">
  96. <picture>
  97. <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/svg?repos=LING71671/open-reverselab&type=Date&theme=dark" />
  98. <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/svg?repos=LING71671/open-reverselab&type=Date" />
  99. <img alt="Star History Chart" src="https://api.star-history.com/svg?repos=LING71671/open-reverselab&type=Date" />
  100. </picture>
  101. </a>
  102. </div>
  103. ---
  104. ## Disclaimer
  105. By accessing or using this project, you agree to be bound by the full disclaimer — covering all versions and branches, all users (direct and indirect), all derivatives, legal compliance across all jurisdictions (including export controls and data protection laws), authorized purposes only, prohibited uses, no warranty, limitation of liability, mandatory disclaimer retention in derivatives, AI/ML training protections, and more.
  106. > :page_facing_up: Full legal text: [DISCLAIMER.md](DISCLAIMER.md) · [中文版](DISCLAIMER.zh.md)
  107. ## License
  108. **Apache-2.0.** See [LICENSE](LICENSE). By contributing, you agree your contributions are licensed under the same terms.

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.