You have access to my Maze workspace through the Maze MCP.
Goal: build a usage dashboard for my Maze team "[TEAM NAME]" for my contract from [START DATE] to [END DATE], as a shareable artifact.

1. Inventory
- Call get_user_details and pick the team named above.
- Run `search` once per study type (entityTypes: unmoderated_study, then moderated_study, then ai_moderated_study) with teamId, query "", includeArchived true, dateFilter { dateField: "createdAt", from, to }. Page with pagination.nextCursor until pagination.complete is true; check each list's row count equals pagination.total.
- Keep: id, name, link, created date, creator name and email, workspace and project (from the location breadcrumb), status.

2. Details per study (split across parallel sub-agents if there are more than ~50 studies; one output file each)
- Unmoderated (get_unmoderated_study):
  - detail "summary": maze type, block counts by block type, recording enabled and required, device restrictions, prototype name, type and provider, and website-test hostnames.
  - detail "sessions" (includeTranscripts false, limit 100, increase offset until hasMore is false): started = all sessions; completed = sessions with a completed status; first and last answer dates; average and median duration of completed sessions; panel sessions.
- Moderated / AI-moderated (get_moderated_study / get_ai_moderated_study, detail "summary"): total, completed and in-progress sessions, first and last session date, average duration where available, languages (AI-moderated).

3. Rules
- A study uses one credit as soon as it has its first response (unmoderated), first participant or session (moderated), or first session (AI-moderated). The credit date is that first answer; for moderated studies use the earliest session and mark it as approximate (≈).
- "Counts towards current contract?" = Yes when the credit date falls inside the contract.
- My allocation is [N] credits. Projection = average daily credits since the start × days in the contract. Show "On track" when the projection is at or below the allocation, otherwise "Over allocation".
- "Level 1" = an unmoderated study with 5+ sessions; its date is when the 5th session came in.

4. Output
- Header with filters: contract to date / last 90 days / last 30 days, plus a workspace selector. All sections follow them, except the credit card, which always shows the full contract.
- Headline figures: studies created (split by type), maze responses (completed), interviews (completed), median of maze medians, mazes that reached Level 1.
- Credit card: used of allocation with what's left, a meter split by study type with a projection marker, the on-track status, a one-line projection sentence and the credit rule.
- Credit burn: a cumulative line from the start of the contract to today, a dashed projection to the end date, an even-pace line to the allocation, and a hover tooltip.
- Studies created per month, stacked by type; block usage (% of mazes using each block type); top 10 creators stacked by type.
- Maze completion: started vs completed, the overall rate, and mazes grouped by completion rate (90%+, 75–89%, below 75%).
- Credits by workspace, stacked by type.
- Two searchable, sortable, paginated tables with a filter on "counts towards contract", each linking to its study:
  - Unmoderated: ID, maze name, project, workspace, creator email, type, status, usage consumed at, counts towards current contract?, created at, first response at, last response at, responses started, responses completed, completion rate, average and median response duration (seconds), reached Level 1 at, enabled recording, required recording, total blocks plus one count column per block type, net credits consumed, panel sessions, prototype name, prototype type, prototype provider, device restrictions, start URL hosts.
  - Moderated & AI-moderated: UUID, is AI moderated, name, status, creator email, created at, usage consumed at, counts towards current contract?, total, completed and in-progress interviews, project, workspace, first and last interview started at, net credits consumed.

Use [our brand colors / a clean neutral style], with one color per study type used consistently. Add a methodology note covering the credit rule, the Level 1 definition, the projection, and known gaps: deleted studies and waived credits aren't visible through the MCP, and moderated credit dates are approximate.