Overview

In Skill Assessment, you can review how the measured members use AI tools, drilling down from the whole organization to a team and then to an individual member.

Screen What you can see
AI Monitoring Overview Organization-wide KPIs, AI activity, AI tool usage, rank distribution, teams, and members
Team details Team score, team KPIs, AI activity, AI tool usage, Skill / MCP usage, rank distribution, and the team's members
Member List Status, usage, and last log of each measured member, with filters and sorting
Member Detail An individual member's KPIs, AI activity, AI tool usage, and Skill / MCP usage

These screens are available to members with Admin Authorization or Manager Authorization. The Skill Assessment section of the sidebar lists AI Monitoring Overview, Member List, AI Readiness, and My Page.

Screens for admins and managers currently do not show individual members' scores or ranks.

AI Monitoring Overview (organization-wide)

Click AI Monitoring Overview in the sidebar.

Use the selector at the top right to choose the aggregation window: Last 7 days, Last 30 days, or Last 90 days (Last 7 days by default). The window you choose is kept when you open a team and come back.

KPI cards

Card What it shows
Organization average score The organization's average score (out of 100) and its change from the previous period
Active rate (weekly) The share and number of active members among the measured members (for example, "8 / 10 members"), and its change from the previous period (pt)
AI tool cost The total cost in the window and the cost per member (in USD)
Token usage The total token usage in the window and the usage per member

AI activity

Aggregated from the usage logs of the whole organization.

  • Total requests and Tokens used: the totals for the window and the values per member
  • Average prompt length: the number of characters of the prompt per request
  • Active members: active members / measured members
  • Usage breakdown (by request ratio): the share of requests by purpose, such as Code Generation, Code Review, Debugging, Refactoring, Documentation, Research, and Test
  • Languages & stacks: the share of programming languages and technology stacks handled in requests
  • Active hours: token usage by hour of the day, with the peak hour and the share of off-hours usage (22:00 onward)

Times are shown in the time zone of the person viewing the screen.

AI tool usage

For each AI tool, the table shows its share of requests, the change in that share from the previous period (pt), cost, cost per user, and number of users. Open a tool's row to see the breakdown by model within that tool.

Below the table, Total cost, Per member, vs. previous period (change in cost), and Tools in use are shown.

Rank distribution

Shows how many measured members are in each rank (S, A, B, C, D) and how many are Not Evaluated. Go to member list opens the list of the members counted here.

Teams

The groups created in your company settings are listed as teams, ordered by usage. For each team, the list shows the number of members and active members, rank, score, and token usage.

  • A team with fewer than two evaluated members does not show a score.
  • If there are no groups, you can create one from Set up groups. For how to create groups, see Manage Members in Groups.

Click a team's row to open its Team details.

Members

The measured members are listed in order of usage. Each row shows token usage and the primary tool (the AI tool the member uses most). Click a row to open the Member Detail.

Team details

From the Teams list on the AI Monitoring Overview, click a team.

The top of the screen shows the team name, the number of members, and the aggregation window. As on the AI Monitoring Overview, you can choose Last 7 days, Last 30 days, or Last 90 days.

Team score

Shows the team's rank and score, and the change from the previous period. It also shows the team's score for each of the five SEI indicators (such as AI Collaboration Skills, Problem Solving, and Communication) and the difference from the organization average.

A team with fewer than two evaluated members does not show a team score, so that an individual member's score cannot be inferred from the team score.

KPI cards

Card What it shows
Active members Active members / team headcount, and the active rate
Total requests The total for the window and the requests per day per member
AI tool cost The total for the window, the cost per member, and the difference from the organization (%)
Token usage The total for the window and the usage per member

The active rate is calculated against the whole team, including members who have not started measurement yet.

AI activity

Shows the team's totals compared with the organization average.

  • Total requests: the value per member and the organization's value per member
  • Tokens used: broken down into input and output
  • Cache hit rate: compared with the organization average
  • Sessions: the number of sessions per member
  • Max concurrent: the maximum number of sessions running at the same time, and the organization's value
  • Usage breakdown (by request ratio) and Languages & stacks (top 4)
  • Active hours: token usage by day of the week and hour, with the peak day and hour and the share of off-hours activity (22:00 onward)

AI tool usage

Shows each AI tool's share of requests and its cost in the window. Open a tool's row to see the breakdown by model.

Skill / MCP usage

Shows the skills and MCP servers most used in the team, split into Top skills and Top MCP servers / plugins, with the number of times each was used. You can switch between AI tools in each list.

Rank distribution

Shows how many of the team's evaluated members are in each rank (S, A, B, C, D) and how many are Not Evaluated, together with the team's average score and the organization's average score. It is not shown for a team with fewer than two evaluated members.

Members

The team's members are listed in order of token usage in the window. Each row shows token usage, cost, primary tool, and last log (the date and time the latest usage log arrived). Click a row to open the Member Detail.

Member List

Click Member List in the sidebar. All members invited for measurement are listed.

Column What it shows
Name Name and email address
Status The measurement status
Primary tool The AI tool the member uses most. Shows Not detected if no usage log has arrived yet
Tokens Used Token usage
Cost AI tool cost
Last log The date and time the latest usage log arrived

If a member's Last log stays old, their measurement may have stopped, for example because their AI tool settings changed.

The status is one of Invitation Sent, Awaiting connection, Connected, Measuring, Connection failed, Measurement failed, or Measurement ended.

Filtering and sorting

  • Search by Name or Email Address: filters by a partial match on name or email address
  • Status, Tokens Used, and Cost: filter by status, and by a range of token usage or cost
  • Sorting: sort by created date, status, token usage, or cost, in ascending or descending order

Ending measurement and deleting a member

Use the menu (︙) on each row.

  • End measurement: stops the member's measurement agent, so their AI activity is no longer recorded. Data and scores already measured are kept. This cannot be undone.
  • Delete: removes the member from the list. Deleting does not stop the measurement, so end the measurement first if you also want to stop it. This cannot be undone.

To add members for measurement, use Invite Member at the top right. For the steps, see Setting Up Skill Assessment.

Member Detail

Click a member's row in the Member List, the AI Monitoring Overview, or Team details.

The top of the screen shows the member's name, email address, status, and registered date. For a member whose status is not Measurement ended, you can end the measurement with End measurement. You can also delete the member from this screen.

The Member Detail shows usage-related items. It does not show the member's score, rank, or score breakdown.

KPI cards

Card What it shows
Active days (30d) The number of days in the last 30 days the member used AI tools
Total requests The number of requests
AI tool cost (last 30 days) The cost in the last 30 days
Token usage (last 30 days) The token usage in the last 30 days

AI activity

Aggregated from the usage logs of the last 30 days.

  • Total requests and Tokens used
  • Cache hit rate
  • Sessions
  • Max concurrent: the maximum number of sessions running at the same time, and the average concurrency
  • Usage breakdown (by request ratio) and Languages & stacks
  • Active hours: token usage by day of the week and hour over the last 7 days, with the peak day and hour and the share of off-hours activity (22:00 onward)

Values that have not been measured yet are shown as "-".

AI tool usage

Shows each AI tool's share of requests and its cost. Open a tool's row to see the breakdown by model.

Skill / MCP usage

Shows the skills and MCP servers most used in the last 30 days, split into Top skills and Top MCP servers / plugins, with the number of times each was used. You can switch between AI tools.

Notes

  • Skill Assessment is in beta. The items shown and how they are aggregated may change.
  • Team-level scores are shown only when two or more members are evaluated.
  • To see how the skills, MCP servers, and repositories in your organization are used, check AI Readiness. For details, see Checking AI Readiness.