🔐

MyAIServer Dashboard

This dashboard is protected.
Request a one-time code, then enter your passphrase to unlock full access.

✅ Code sent to your email — valid for 10 minutes
Invalid code or passphrase. Please try again.

MyAIServer — Security Dashboard

MyAI-IT · AI System & Security Administrator

204.168.194.138
Audit: 2026-04-10 · 22:58 EDT
🔴 0 Critical
🟠 1 High
⚠️ 18 Medium
⏰ 2 Recurring
🖥️

Server Overview

Hostname
MyAIServer
Kernel
6.8.0-107-generic
Uptime
up 21 hours, 5 minutes
Disk /
301G 14G 275G 5%
Memory total/used/free
15Gi 1.5Gi 3.7Gi
Load average
0.09 0.04 0.00
OpenClaw Version
Runtime-reported current version
2026.4.9
⚙️

Services

OpenClaw Gateway
systemd user unit state from this run
● Inactive
nginx
reverse proxy / web edge
● Active
Recurring Tasks
Live scheduler inventory recalculated on every run
2 recurring jobs
Gateway Bind
runtime config inspection
loopback
📡

OpenClaw / Channels

✈️
Telegram
default + myai-it accounts configured
configured
💬
WhatsApp
channel enabled in runtime config
enabled
🔐
Gateway Auth
token mode with loopback bind
hardened
🧰
FS Sandbox Policy
workspaceOnly in config, but trusted shell access remains powerful
review
🧪

Sanitized Findings

🔴

Sanitized markdown secret exposure

A workspace markdown file contains a raw deployment credential reference and should be cleaned immediately. Report sanitized; secret value omitted.

🟡

Instruction drift across agent guidance

Approval wording conflicts exist between files, including approve vs approved/confirm, which can cause unsafe or inconsistent execution behavior.

🟡

Group-writable markdown instruction files

14 markdown files under agent/workspace scope are mode 664 and should be reduced to 640 or 600 where appropriate.

🟡

Gateway service state mismatch

openclaw-gateway user service reports inactive while the environment remains reachable enough for this cron run. Service install/state should be reconciled.

Recurring Tasks

myai-it-dashboard-scan
Owner: myai-it · Creator: myai-it · every 6h · Telegram announce to this chat
active
AI Morning Brief
Owner: main · Creator: main · cron 0 6 * * * America/New_York · WhatsApp announce to +972547777577
active
📁

Markdown Scan Summary

Secret leakage findings
sanitized only, no values exposed
1
Instruction drift findings
conflicting approval / policy wording
4
Safety tone findings
subordinate identity wording worth review
2
Markdown permission findings
group/world writable markdown files
14
⚠️

Pending Actions

1

Sanitize workspace memory and deploy references

Remove raw Cloudflare token material from markdown memory/history files and replace with a note that secrets live only in secure environment storage.

2

Unify approval policy text

Normalize all agent/workspace instruction files to one exact approval rule to eliminate drift and automation ambiguity.

3

Tighten markdown file permissions

Change agent/workspace markdown guidance files from group-writable 664 to a stricter mode such as 640 or 600.

4

Reconcile gateway service state

Review the user unit and startup path for openclaw-gateway because systemd reports inactive despite the gateway environment being usable.

5

nginx global TLS cleanup still pending

Global nginx config still needs deprecated TLS protocol entries removed and server_tokens disabled during an approved maintenance window.

6

Firewall verification still pending

UFW/iptables/nftables status remains unconfirmed from this run. Verify host or provider firewall ownership explicitly.

7

Self-signed TLS still in use

If a public domain is available, replace the current self-signed certificate with a managed certificate.