CRITICAL fixes: - Path traversal prevention in brain/backup/skill/file endpoints - Unsafe tar extraction with safe_extractall() path validation HIGH fixes: - XSS: encodeURIComponent for onclick handlers in skills/memory/prompts/standards/backups - XSS: escapeHtml() for user-controlled text in cost/plugins/settings/audit/dashboard - XSS: safe class name mapping in setup-wizard (attribute injection) - Security headers middleware (CSP, HSTS, X-Frame-Options, X-Content-Type-Options) - CORS restricted to localhost:8080 - Settings API keys masked before returning to client - Session replay content limited to 2000 chars + path validation MEDIUM fixes: - Chat message length limit (10000 chars) - Input validation on brain/skill/backup file names (reject .. and /) |
||
|---|---|---|
| .. | ||
| pages | ||
| api.js | ||
| app.js | ||
| index.html | ||
| styles.css | ||
| utils.js | ||