Dashboard Features

The Clawkeeper web dashboard at clawkeeper.dev/dashboard provides fleet monitoring, security intelligence, and alerting for your Claude Code workstations and AI agent deployments.

Overview

The dashboard organizes your security operations into two primary views — Claude Code workstations and AI agents — plus shared monitoring and intelligence tools:

Workstations — fleet management for Claude Code developer workstations. Online/offline status, version tracking, session monitoring, and project visibility. See Workstations.
Claude Code Security — unified security dashboard with Runtime Shield stats, threat patterns, MCP skill inventory, and policy enforcement status. See Runtime Shield.
Activity — unified audit trail of all Claude Code tool calls and AI agent events. Server-side pagination, filtering by verdict/severity/host, full-text search.
Fleet overview — see all hosts, their grades, scores, and last scan times at a glance
Host detail — Claude Code workstations show online status, sessions, developer info, and projects. Cloud-hosted AI agents show scan results, shield events, and a Manage tab for gateway configuration.
Security Insights — split into Posture (scan insights: CVEs, credentials, drift, misconfigs) and Attacks (shield events: threat stats, top patterns, timeline). Sortable and filterable by host, severity, and time.
Activity stream — chronological feed of scans, grade changes, check flips, and host registrations
AI-powered insights — automated security analysis including skills supply chain and integrity drift (Pro+)
CVE intelligence — live vulnerability feed cross-referenced against your installations (Pro+)
Cloud Deploy — provision managed OpenClaw instances in under 60 seconds with golden images (Pro+)
Console — browser-based terminal to cloud instances via Tailscale mesh (Pro+)
Gateway management — edit agent instructions, manage skills, and configure channels from the dashboard without SSH (Pro+)
Skill Marketplace — browse, search, and install community skills from ClawHub's 13,000+ skill catalog with security scoring (Pro+)
Email alerts — get notified when issues are detected (Pro+)
Webhook alerts — integrate with Slack, PagerDuty, and custom endpoints (Team+)

Cloud instances are available on paid plans (Pro, Team, and Enterprise) only. Free accounts cannot deploy or access cloud instances.

Free tier

Every dashboard account includes:

Full 55-check CLI scanner with auto-fix
1 Claude Code workstation with Runtime Shield
1 connected GitHub repo
1 host on the dashboard
7-day scan history
1 API key
Letter grade and score tracking

Pro features

Upgrade to Pro (from $19/mo) to unlock:

CVE vulnerability audit

The dashboard cross-references your scan results with the live OpenClaw CVE feed. When a scan detects CVEs affecting your installed version, the dashboard creates insights with:

CVE ID and link to the NVD entry
Severity level and CVSS score
Affected packages and version ranges
Specific upgrade guidance

AI-powered insights

The insight engine analyzes every scan and generates actionable findings:

Insight Type	Description
Critical failure	High-severity checks that failed (privileged mode, credential exposure, container escape)
Credential exposure	Secrets found in configs, shell history, MEMORY.md, or session logs
Prompt injection	Injection patterns detected in sessions, memory, or skills
Skill supply chain	Dangerous install commands, secret injection, exfiltration patterns, or prompt-level attacks in skills
Skill integrity drift	Skill files changed since last recorded hash — possible tampering
CVE vulnerability	Known CVEs matching your installed OpenClaw version
New regression	Checks that previously passed but now fail
Grade degradation	Host grade dropped since last scan
Fleet inconsistency	Same check passes on some hosts but fails on others
Stale host	Host hasn't reported a scan in over 7 days
Quick win	Easy-to-fix checks with specific remediation steps

Each insight includes severity, affected hosts, and step-by-step remediation instructions.

Score trends

Visualize how your security posture changes over time with score history charts on each host detail page. Spot regressions early before they become critical.

Cloud Deploy

Deploy fully managed OpenClaw instances from the dashboard. Each instance is a dedicated VM with OpenClaw, the Clawkeeper agent, Runtime Shield, and a Tailscale node pre-installed. The wizard walks you through region, compute plan, LLM provider, and security settings. See Cloud Deploy for the full reference.

Console

Click Open Console on any cloud instance to get a browser-based terminal. The connection routes through a Tailscale mesh with JWT authentication — no SSH ports, no VPN, no firewall rules. See Console for the security model.

Gateway management

The Manage tab on cloud host detail pages lets you configure your OpenClaw gateway without SSH access:

Agent Instructions — edit the SOUL.md that defines your agent's personality and behavior
Skills — browse, add, and remove bundled skills from the gateway's skill catalog
Skill Marketplace — search and install community skills from ClawHub directly from the dashboard. Each skill shows a security badge (Clean, Suspicious, or Blocked) based on ClawHub's automated scanning. Skills are cached locally and synced every 6 hours.
Channels — add Telegram, Discord, or WhatsApp channels to connect your agent to messaging platforms

Changes are applied in real time via WebSocket RPC through the Tailscale mesh.

Email & webhook alerts

Configure notifications to alert you when Clawkeeper detects issues. Email alerts are available on Pro and above. Webhook alerts require a Team or Enterprise plan. See Notifications for setup details.

Extended history

Pro accounts retain 90 days of scan history, Team and Enterprise accounts have unlimited retention (vs 7 days on free).

Getting started

Claude Code (recommended)

Create an account at clawkeeper.dev
Generate an API key in Settings
Add hooks to ~/.claude/settings.json — see Claude Code Setup
Start a Claude Code session — your workstation appears in the dashboard automatically

CLI Scanner

Create an account at clawkeeper.dev
Generate an API key in Settings
Install the agent on your hosts:

export CLAWKEEPER_API_KEY="ck_live_..."
clawkeeper.sh agent --install

View results at clawkeeper.dev/dashboard
Or deploy a cloud instance for a fully managed experience (Pro+)

See Configuration for agent setup details.