Dashboard and Statistics

The current version of MCP2Skill provides two observation perspectives:

• Dashboard: better for quickly checking the overall status
• Statistics: better for reviewing performance over a time range

These capabilities matter because most AI clients are almost a black box when it comes to MCP. You usually cannot see:

• Which MCP is called most frequently
• Which tool has a higher error rate
• Which client is making high-frequency calls
• Whether your configuration changes have actually taken effect

The dashboard and statistics pages in MCP2Skill are designed to turn these blind spots into observable information.

What the Dashboard Is Best For

The dashboard primarily answers "how are things overall right now" and "where should I look first."

From here you can usually see at a glance:

• Number of installed services
• Number of currently running services
• Number of available tools
• Number of calls today
• Success rate
• Average response time
• Recent call records

If you have just finished a configuration change, it is a good idea to check the dashboard first to confirm everything is stable overall.

Key Areas of the Dashboard

Overview Metrics

Good for quickly determining:

• Whether there have been any actual calls today
• Whether current latency and success rate are normal
• Whether the total number of services and their online status match expectations

Activity Rhythm and Heatmap

Good for observing:

• Which time periods are the most active
• Whether there have been obvious gaps or spikes in recent days

Service Load

Good for identifying:

• Which services handle the most calls
• Whether any single service is clearly carrying a disproportionate load

Client Leaderboard

Good for determining:

• Who is primarily calling MCP2Skill
• Which client is the most active
• Whether a particular client's success rate or response time is abnormal

Recent Calls

Good for quick spot checks, such as:

• Where the most recent call came from
• Whether a failure just happened
• Whether it is related to a change you just made

Key Questions These Two Pages Can Answer

• Which MCP has been the most active recently
• Which client is calling MCP2Skill at high frequency
• Which tool might be the least stable
• Whether call volume and success rate changed after modifying a workspace or service configuration
• Whether the system as a whole became more or less stable after an integration or change

What the Statistics Page Is Best For

The statistics page emphasizes "analysis over a time range."

Common time ranges in the current version include:

• Today
• Yesterday
• Last 7 days
• Last 30 days

When you want to answer questions like "has the last week been more stable," the statistics page is more helpful than the dashboard.

Common Charts on the Statistics Page

They typically include:

• Call trends
• Success vs. failure comparison
• Average response time
• Statistics by service
• Statistics by client
• Top tools

You do not need to memorize the technical details of every chart. Just keep one principle in mind:

When to Focus on Statistics

Go straight to the statistics page in the following situations:

• You want to compare yesterday's performance with today's
• You suspect a service's error rate has risen recently
• You want to know which client makes the most calls
• You want to confirm whether a tool is really being used at high frequency

Usage Recommendations

• Start with the dashboard for routine checks
• Move to the statistics page when investigating trend changes
• After spotting an anomaly, follow up with the logs page

Related Chapters

• To see specific failure details: see Logs and Diagnostics
• To handle issues with a service itself: see MCP Service Management
• To troubleshoot client configuration issues: see Connect External AI Clients