Security
How ApiArk protects your data and our approach to security.
Reporting Vulnerabilities
- -Email security@apiark.dev with details of the vulnerability.
- -You can also report via GitHub Security Advisories on our repository.
- -We will acknowledge receipt within 48 hours and provide a timeline for a fix.
- -Please do not disclose publicly until we have released a patch.
Architecture Security
- -Tauri v2 capability-based permissions — each window gets only the IPC commands it needs.
- -Content Security Policy: no eval(), no remote scripts, no inline scripts.
- -Frontend cannot execute arbitrary shell commands or access filesystem outside scoped directories.
- -All IPC commands are explicitly defined — no open-ended shell access from the webview.
Secret Handling
- -Environment secrets from .env files are loaded into Rust memory only.
- -Secrets are never written to SQLite history — auth tokens are replaced with [REDACTED].
- -Secrets are never written to log files.
- -Secret variables are masked in the UI by default, revealable per-variable.
Data Safety
- -All file writes use atomic write: write to .tmp file, then rename() to target.
- -File saves are serialized per-collection to prevent race conditions.
- -SQLite uses WAL mode for crash resilience.
- -On startup, SQLite integrity is checked — corrupt databases are backed up and a fresh DB is created.
Dependency Auditing
- -cargo audit runs in CI on every pull request, blocking merge on known vulnerabilities.
- -pnpm audit runs in CI, blocking on high/critical severity.
- -New dependencies are manually reviewed before adding: license, maintenance, security history.
Network Security
- -ApiArk makes no network calls by default — zero telemetry, zero analytics.
- -License validation (Pro/Team only) sends license key + app version, nothing else.
- -Update checks are non-blocking and silently skipped if offline.
- -Proxy capture mode binds to 127.0.0.1 only — not accessible from the network.
Last updated: March 2026