-
-
Notifications
You must be signed in to change notification settings - Fork 0
Security
🌍 English translation is auto-generated on release. See the French version for the current complete page.
WorkPilot AI's three-layer security model: OS sandbox, filesystem restriction, dynamic allowlist. Plus credentials storage in the OS keychain.
WorkPilot AI's security model is built around isolation and least privilege, since agents execute code and shell commands autonomously on the user's behalf.
Three-layer security model:
| Layer | Purpose |
|---|---|
| OS Sandbox | Bash commands run in isolation |
| Filesystem Restrictions | Operations are limited to the project directory |
| Dynamic Command Allowlist | Only commands approved for the detected project stack are permitted |
Workspace isolation. Every task runs in a dedicated Git worktree, so the agent pipeline (Planner → Coder → QA Reviewer → QA Fixer) never touches the main branch directly. Changes only reach main through an AI-powered semantic merge once a human reviews and approves the work.
Credential handling. Authentication credentials — whether an OAuth token from a Claude Code subscription or an API key for another provider — are stored in the OS-native secure store (Keychain on macOS, Credential Manager on Windows). The multi-profile system supports automatic token refresh and can switch between registered accounts when one hits a rate limit, avoiding the need to keep raw credentials in plaintext config.
Release integrity. Every published release ships with SHA256 checksums and VirusTotal scan results, giving users a way to verify a downloaded binary hasn't been tampered with before installation.
Team deployments. In Multi-User Server Mode, authentication is handled via JWT or Microsoft Entra with per-user claims, and self-service signup is invitation-only with rate-limited SMTP delivery. Outbound notifications (Slack, email, webhook) include SSRF hardening to prevent them from being abused to reach internal network resources.
For details on supported versions and how to report a vulnerability, see the project's Security Policy.
➡️ Continue: Customization
Getting started / Pour débuter
- 🏠 Home
- 📘 Introduction · fr
- ⚡ Installation · fr
- 🚀 First project · fr
- 🧠 Key concepts · fr
- ❓ FAQ · fr
Usage
- 🖥 User interface · fr
- 📝 Creating a spec · fr
- 🔁 Multi-agent pipeline · fr
- 🤖 Specialized agents · fr
- 🔌 Integrations · fr
- 💡 AI providers · fr
- 🧩 Memory system · fr
Advanced / Avancé
Community / Communauté