Skip to content
Back to search

claudit-sec

Security audit tool for Claude Desktop and Claude Code on macOS — single-command visibility into MCP servers, extensions, plugins, connectors, scheduled tasks, and permissions.

Safety scan

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

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/harmonicsecurity--claudit-sec/badge.svg)](/listings/harmonicsecurity--claudit-sec)

HTML

<a href="/listings/harmonicsecurity--claudit-sec"><img src="/api/listings/harmonicsecurity--claudit-sec/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.

Prerequisites

**macOS:**

| Requirement | How to check | How to install | |-------------|-------------|---------------| | 🍎 **macOS** | You're on a Mac | — | | 🐚 **zsh** | `zsh --version` | Ships with macOS since Catalina | | 🔧 **jq** | `jq --version` | `brew install jq` |

**Windows:**

| Requirement | How to check | How to install | |-------------|-------------|---------------| | 🪟 **Windows 10/11** | You're on a PC | — | | ⚡ **PowerShell 5.1+** | `$PSVersionTable.PSVersion` | Ships with Windows 10+ | | — | No additional dependencies | Fully self-contained |

Install & Run

**macOS:**

git clone https://github.com/HarmonicSecurity/claudit-sec.git
cd claudit-sec
chmod +x claude_audit.sh
./claude_audit.sh

**Windows:**

git clone https://github.com/HarmonicSecurity/claudit-sec.git
cd claudit-sec
powershell -ExecutionPolicy Bypass -File claude_audit.ps1

That's it. The script reads your Claude configuration and prints a colour-coded report to the terminal. It never modifies anything.

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 · 219 lines

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

  1. # 🛡️ CLAUDIT-SEC
  2. **Security audit tool for Claude Desktop on macOS and Windows — including CoWork, extensions, plugins, MCP servers, connectors, and scheduled tasks.**
  3. One command. Full visibility. Read-only.
  4. > ⚠️ **Windows support is a work in progress.** We're aware of a few kinks and bugs and wanted to get something out sooner rather than later. Community feedback and contributions are welcome.
  5. <p align="center">
  6. <img src="media/claudit-terminal.png" alt="CLAUDIT terminal output" width="700">
  7. </p>
  8. ## 🤔 Why
  9. Claude Desktop introduces a new class of endpoint risk: AI agents with autonomous execution, persistent scheduled tasks, MCP server integrations, browser-control extensions, and OAuth-authenticated connectors to external services. Most of this configuration lives in JSON files scattered across multiple directories with no centralised visibility.
  10. CLAUDIT gives you that visibility in a single command.
  11. > 📝 **A note on "Code":** Claude Desktop includes a built-in agent coding feature called **Code** (visible in the app's sidebar). This is **not** the same as **Claude Code**, the standalone terminal CLI. CLAUDIT primarily audits Claude Desktop and its CoWork features. It does include a basic check of the Claude Code settings file (`~/.claude/settings.json` on macOS, `%USERPROFILE%\.claude\settings.json` on Windows), but the focus is squarely on the Desktop app.
  12. ## 📋 What It Audits
  13. | Area | What's Checked |
  14. |------|---------------|
  15. | 🖥️ **Desktop Settings** | `keepAwakeEnabled`, sidebar/menuBar preferences |
  16. | 🤖 **CoWork Settings** | Scheduled tasks, web search, browser use, dispatch (mobile→desktop), network mode, egress policy, enabled plugins, marketplaces |
  17. | 🏢 **Workspaces** | Multi-workspace detection, account names, session counts, org indicators (DXT-managed, org-plugins, dispatch-bridge) |
  18. | 🔌 **MCP Servers** | Server names, commands, arguments, environment variable keys |
  19. | 🧩 **Extensions (DXT)** | Installed extensions, signature status, dangerous tool grants |
  20. | ⚙️ **Extension Settings** | Per-extension allowed directories and configuration |
  21. | 🚦 **Extension Governance** | Allowlist enabled/disabled, blocklist entries |
  22. | 📦 **Plugins** | Installed, remote (org-deployed), cached (downloaded) |
  23. | 🪝 **Plugin Hooks** | Lifecycle hooks executing shell commands (PreToolUse, PostToolUse, Stop, etc.) |
  24. | 🔗 **Connectors** | OAuth-authenticated web services, desktop integrations |
  25. | 🎯 **Skills** | User-created, scheduled, session-local, and plugin skills across 9 paths |
  26. | ⏰ **Scheduled Tasks** | Task names, cron expressions (with plain English translation) |
  27. | 🔐 **App Config** | Network mode, extension allowlist/blocklist keys, device identifiers |
  28. | 📲 **Dispatch** | Bridge state (OFF/CONFIGURED/ON), active session detection via `hostLoopMode` and `bridge-state.json` |
  29. | 🔇 **Disabled MCP Tools** | Per-session tools explicitly disabled (with dangerous tool callout) |
  30. | 🏃 **Runtime State** | Running processes, sleep assertions, LaunchAgents, crontab entries |
  31. | 🍪 **Cookies** | `Cookies` and `Cookies-journal` presence |
  32. > 📖 For a detailed breakdown of every individual check, what it means, and why it matters, see the [Findings Reference](docs/findings-reference.md).
  33. ## ⚡ Getting Started
  34. ### Prerequisites
  35. **macOS:**
  36. | Requirement | How to check | How to install |
  37. |-------------|-------------|---------------|
  38. | 🍎 **macOS** | You're on a Mac | — |
  39. | 🐚 **zsh** | `zsh --version` | Ships with macOS since Catalina |
  40. | 🔧 **jq** | `jq --version` | `brew install jq` |
  41. **Windows:**
  42. | Requirement | How to check | How to install |
  43. |-------------|-------------|---------------|
  44. | 🪟 **Windows 10/11** | You're on a PC | — |
  45. | ⚡ **PowerShell 5.1+** | `$PSVersionTable.PSVersion` | Ships with Windows 10+ |
  46. | — | No additional dependencies | Fully self-contained |
  47. ### Install & Run
  48. **macOS:**
  49. ```bash
  50. git clone https://github.com/HarmonicSecurity/claudit-sec.git
  51. cd claudit-sec
  52. chmod +x claude_audit.sh
  53. ./claude_audit.sh
  54. ```
  55. **Windows:**
  56. ```powershell
  57. git clone https://github.com/HarmonicSecurity/claudit-sec.git
  58. cd claudit-sec
  59. powershell -ExecutionPolicy Bypass -File claude_audit.ps1
  60. ```
  61. That's it. The script reads your Claude configuration and prints a colour-coded report to the terminal. It never modifies anything.
  62. ## 🎛️ Usage
  63. **macOS:**
  64. ```
  65. ./claude_audit.sh [OPTIONS]
  66. Options:
  67. --html [FILE] Generate a standalone HTML report
  68. --json Output structured JSON
  69. --user USER Audit a specific user
  70. --all-users Audit all users with Claude data (requires root)
  71. -q, --quiet Only show WARN and CRITICAL findings
  72. --version Print version and exit
  73. -h, --help Show usage
  74. ```
  75. **Windows:**
  76. ```
  77. powershell -ExecutionPolicy Bypass -File claude_audit.ps1 [OPTIONS]
  78. Options:
  79. -Html [FILE] Generate a standalone HTML report
  80. -Json Output structured JSON
  81. -User USER Audit a specific user
  82. -AllUsers Audit all users with Claude data (requires admin)
  83. -Quiet Only show WARN and CRITICAL findings
  84. -Version Print version and exit
  85. -Help Show usage
  86. ```
  87. ### Examples
  88. **macOS:**
  89. ```bash
  90. # Default: colour output in terminal
  91. ./claude_audit.sh
  92. # Only warnings and critical findings
  93. ./claude_audit.sh -q
  94. # Standalone HTML report
  95. ./claude_audit.sh --html
  96. # JSON for SIEM ingestion
  97. ./claude_audit.sh --json > audit.json
  98. # Specific user
  99. ./claude_audit.sh --user jsmith
  100. # All users (run as root via MDM — FleetDM, Jamf, Mosyle, CrowdStrike RTR)
  101. sudo ./claude_audit.sh
  102. ```
  103. **Windows:**
  104. ```powershell
  105. # Default: colour output in terminal
  106. powershell -ExecutionPolicy Bypass -File claude_audit.ps1
  107. # Only warnings and critical findings
  108. powershell -ExecutionPolicy Bypass -File claude_audit.ps1 -Quiet
  109. # Standalone HTML report
  110. powershell -ExecutionPolicy Bypass -File claude_audit.ps1 -Html
  111. # JSON for SIEM ingestion
  112. powershell -ExecutionPolicy Bypass -File claude_audit.ps1 -Json > audit.json
  113. # Specific user
  114. powershell -ExecutionPolicy Bypass -File claude_audit.ps1 -User jsmith
  115. # All users (run as admin via MDM — Intune, CrowdStrike RTR)
  116. powershell -ExecutionPolicy Bypass -File claude_audit.ps1 -AllUsers
  117. ```
  118. > 💡 **macOS:** When run as **root** (uid 0), the script automatically discovers and scans all users with Claude data. No flags needed.
  119. >
  120. > 💡 **Windows:** When run as **Administrator**, the script can scan all users with the `-AllUsers` flag. MDM tools like Intune and CrowdStrike RTR execute scripts with elevated privileges.
  121. ## 📊 Output Formats
  122. ### 🖥️ Terminal (default)
  123. Colour-coded output with Unicode tables and severity indicators.
  124. <p align="center">
  125. <img src="media/claudit-terminal.png" alt="CLAUDIT terminal output" width="700">
  126. </p>
  127. ### 🌐 HTML (`--html`)
  128. Standalone dark-themed report with collapsible sections. Created with restrictive file permissions (`0600`).
  129. <p align="center">
  130. <img src="media/claudit-html.png" alt="CLAUDIT HTML report" width="700">
  131. </p>
  132. ### 📄 JSON (`--json`)
  133. Structured output for SIEM ingestion. Sensitive fields (OAuth tokens, API keys, secrets) are automatically redacted. Multi-user scans produce a JSON array.
  134. ## 🚨 Severity Levels
  135. | Severity | Meaning |
  136. |----------|---------|
  137. | 🟠 **WARN** | Increases risk surface — e.g. unsigned extensions, autonomous execution enabled |
  138. | 🟡 **REVIEW** | Needs human judgement — e.g. org-deployed plugins, MCP servers present |
  139. | 🔵 **INFO** | Informational — e.g. Claude is running, permissions granted |
  140. ## 📖 Documentation
  141. | Doc | Description |
  142. |-----|-------------|
  143. | [Findings Reference](docs/findings-reference.md) | Every individual check CLAUDIT performs, what it means, why it matters (risk, compliance, AI enablement), and what to do about it |
  144. ## 🔒 Security Properties
  145. - **Read-only** — never writes to, modifies, or deletes any audited file
  146. - **No network access** — all data collected from local filesystem and system commands
  147. - **Sensitive data redacted** — tokens, keys, and secrets replaced with `[REDACTED]` in all output formats
  148. - **Minimal privileges** — runs as current user; root only needed for multi-user scans
  149. - **Single file** — macOS requires `jq`; Windows is fully self-contained (no external dependencies)
  150. - **Auditable** — the entire tool is one readable script (`claude_audit.sh` on macOS, `claude_audit.ps1` on Windows)
  151. ## 💜 Built with Claude Code
  152. This project is built and maintained using [Claude Code](https://docs.anthropic.com/en/docs/claude-code). We love it. Seriously. If you're building developer tools and haven't tried it yet, you're missing out.
  153. ## 📄 License
  154. Apache License 2.0 — see [LICENSE](LICENSE) for details.

Ratings and comments

Sign in to rate this skill.

Sign in

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