Reports

Reports provides structured training analytics on TA performance across your institution. It computes overview tables, leaderboard rankings, trend charts, and detailed attempt history — all filterable by date range, cohort, department, simulation, and scenario.

What is Reports?

Reports is the analytics engine that aggregates TA training data into five computed sections:

• Header Metrics — Total attempts, average score, completion percentage, first-attempt pass rate (each with trend data and hover tooltips).
• Overview — Per-simulation breakdown showing attempts, completed attempts, passed attempts, average score, completion percentage, and pass rate.
• Leaderboard — Ranked TA profiles with metrics like average score, highest score, total attempts, and first-attempt pass rate.
• Trends — Time-series chart data with daily aggregations of attempts, scores, and pass rates.
• History — Paginated attempt rows with scores, chat counts, and completion details.

Reports is designed for program coordinators who need to evaluate how a cohort of TAs is performing across Confused, Aggressive, Passive, and Happy student persona simulations.

Quick Start

CLI

Calls below use $GLOW_INSTANCE_URL + $GLOW_TOKEN — see Authentication to export them once.

# Fetch the full reports bundle
glow attempts report
 
# Fetch reports filtered by cohort and date range
glow attempts report --body '{"cohort_ids": ["fall-2025-cs101"], "start_date": "2025-01-01", "end_date": "2025-06-30"}'
 
# Export all reports data as a ZIP
glow attempts export

API

# Get reports with filters
curl -X POST $GLOW_INSTANCE_URL/attempt/report \
  -H "Authorization: Bearer $GLOW_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "department_ids": ["dept-cs"],
    "simulation_ids": ["sim-confused-student"],
    "start_date": "2025-01-01",
    "end_date": "2025-06-30",
    "sort_by": "average_score",
    "sort_order": "desc",
    "page_limit": 50
  }'

Filtering

Understanding the Response

The ReportsResponse contains:

sections

The computed report sections:

• header_metrics — total_attempts, average_score, completion_percentage, first_attempt_pass_rate. Each metric includes current_value, has_data, data_points, and hover details (mean, median, mode, count, percent).
• overview — Array of ReportsOverviewRow grouped by simulation: simulation_id, attempts, completed_attempts, passed_attempts, average_score, pass_rate.

• leaderboard — Array of ReportsLeaderboardRow with rank, profile_id, total_attempts, average_score, highest_score, plus detailed profile_metrics (messages per session, session efficiency, stagnation rate, time spent).

• trends — Array of ReportsTrendPoint with date, attempts, average_score, pass_rate for charting.

• history — Array of ReportsHistoryRow with attempt_id, score_percent, has_passed, num_chats, total_time_seconds.

resources

Resource metadata keyed by ID for hydrating IDs into display names:

• simulations, profiles, scenarios, cohorts, personas, rubrics

Filter options

• simulation_options, profile_options, scenario_options — dropdown options with value, label, and count.

Refreshing and Exporting

# Refresh materialized views and caches
curl -X POST $GLOW_INSTANCE_URL/attempt/refresh \
  -H "Authorization: Bearer $GLOW_TOKEN"
 
# Export all reports data as a denormalized ZIP
curl -X POST $GLOW_INSTANCE_URL/attempt/export \
  -H "Authorization: Bearer $GLOW_TOKEN"

The export returns content (base64-encoded), file_name, mime_type, and row_count.

Common Operations

Related

• Reports API Reference
• Reports CLI Reference
• Dashboard Guide — admin analytics hub
• Leaderboard Guide — TA performance rankings