Inside the dashboard.
Frankie reads every chat your team has on Claude and ChatGPT, scores it against a rubric you control, and surfaces it three ways. One tool. One rubric. Three views.
What an employee sees.
Hover or click for details. Tap for details.
Literacy score. Your overall grade. Banded as Practicing, Solid, or Strong so the headline reads at a glance. Weekly delta and the next band sit beside it.
Activity. Chats, messages, and est. tokens for the rolling window. Tells you whether the score has a real sample. Or three chats from last week.
Pack tabs. Which rubric the score uses. ALL is the union of every pack. The named tabs are starter packs. Custom packs pin here too.
Scores. Weakest and strongest skills with weekly deltas. The gap between them is the point. It says where the next coaching nudge will land.
Recommendations. Today's coaching focus, drawn from yesterday's chats. One specific change to try. The reasoning behind it. The skills it touches.
Trajectory. Score over the last 30 days. Most-improved area called out. Current level beside the baseline. A flat line is not failure. A downward line is a signal.
What a manager sees.
Hover or click for details. Tap for details.
Team literacy score. Team average, banded the same way as an individual. The next band's distance sits beside it, so progress is visible without context.
Members and activity. How many people. How many chats. How much usage. Confirms the team score sits on enough material to mean something.
Pack tabs. Which rubric the team is scored against. Swap packs per team. The dashboard rolls up with whatever pack is active.
Team members. Top performer and team average side by side. The distribution strip below shows the lift, and where the outliers sit.
Signals. Active patterns Frankie has flagged. Each signal carries a severity and a delivery status. You see the ones to act on. Before they hit the work.
Trends. Team trajectory over the last 30 days. Most-improved area called out. Current level beside the baseline. Direction matters more than the absolute number.
A rubric that fits your team, not
a generic one.
Packs are fully customizable. Tune the dimensions, swap them per role, or build your own from scratch. Below are the four starter packs we ship today.
Universal
- Prompt clarity
- LLM deficiency awareness
- Content evaluation
- Outcome success rate
- Learning trajectory
- Multi-model fluency
- Cost awareness
- Responsible use
+ 5 more
Coding
- Verification
- Error recovery
- Agentic delegation
- Tool selection
- Context provision via files
- Iteration style
Marketing
- End goal specification
- Tone, audience, format
- Output evaluation
- Iterative refinement
- Brand context
Research
- Source provision
- Source verification
- Cross-source synthesis
- Probe discipline
- Breadth and depth balance
We take privacy seriously.
We write our privacy posture in plain language. What we collect, what we never do, our sub-processors, and our retention windows are all documented and reviewable today. Read more about how we handle your data.