Reports

The three report views in Elixion, the CSV export, and the report endpoints behind them

Reports (/reports), under Insights in the sidebar, gives you three views of one project and a CSV of whichever one you are looking at. It is the fastest way to get a number out of Elixion, the platform for Engineering Intelligence™, and into a document someone else reads.

Pick a project at the top, then one of three views.

Velocity

Only shown for projects that run sprints.

A bar chart of completed against planned points, sprint by sprint, beside a summary reading:

  • Average Velocity, in story points per sprint, averaged over all completed sprints. This is the same definition the Velocity tab on Charts uses, so the two agree.
  • Sprints Analyzed, the number of completed sprints behind that average.
  • Total Completed, in points.

Read Sprints Analyzed before you read Average Velocity. An average over two sprints predicts nothing.

Distribution

Three breakdowns of the project's issues, each with counts and percentages:

  • By Status
  • By Priority
  • By Type

This is the view for answering "what is this backlog actually made of". A project whose By Type chart is two thirds bugs has a quality problem, not a planning problem.

Overview

Four counts across the top: Total Issues, Completed, In Progress and Total Story Points.

Below them, Issues closed in this period, a percentage with its denominator stated: issues created in the reporting window that are now Done, project wide, by issue count.

That heading is precise on purpose. It is not a sprint figure and it is not story points, so it will not match a sprint's own completion percentage and is not meant to. A sprint's completion figures are on that sprint's Reporting tab.

Exporting

The Export button downloads a CSV of the view you are on.

| View | What the CSV contains | |------|----------------------| | Velocity | One row per sprint with completed and planned points, then the average velocity | | Distribution | Status, priority and type breakdowns, each with counts and percentages | | Overview | The overview metrics as name and value pairs |

Report endpoints

Beyond the page, Elixion exposes reports over the API, for teams pulling figures into their own tooling. All are under /api/v1/reports:

| Endpoint | What it returns | |----------|-----------------| | executive-summary | A cross project summary | | sprint-performance | Sprint metrics for a project and date range | | agent-performance | Agent analytics for a project and date range | | project-health | Health metrics per project | | team-analytics | Team metrics | | automation-impact | What automation has changed | | realtime | The current state | | custom (POST) | A report built from a request body |

Two of these are worth knowing about even if you never call them by hand. Agent performance and automation impact are the reports that answer what the agents are contributing, which no other single surface reports in one place.

Export endpoints

GET /api/v1/reports/export/csv/{report_type} returns CSV for sprint-performance, agent-performance and project-health. Any other type is refused.

GET /api/v1/reports/export/pdf/{report_type} returns a PDF for executive-summary. Any other type is refused.

Both take project_id, date_from and date_to as query parameters. Omit project_id and the report is scoped to the projects you can reach, not to everything.

Example:

curl -X GET \
  "https://api.elixion.ai/api/v1/reports/export/csv/sprint-performance?project_id=<id>&date_from=2026-01-01&date_to=2026-03-31" \
  -H "Authorization: Bearer $ELIXION_TOKEN" \
  -o sprint-performance.csv

Report delivery on a schedule is not part of Elixion today. What ships is the page above, its CSV export, and these endpoints.

Making a report worth reading

  • Say what the denominator is. Every percentage here has one, and a percentage quoted without it is the most common way a status update becomes an argument.
  • Pair velocity with its spread. The Velocity tab on Charts reports a standard deviation; an average without it reads as more certain than it is.
  • Use Distribution before planning, not after. Knowing the backlog is half tech debt changes what you commit to.
  • Export at the same point in a cycle each time. Comparing a mid sprint export to an end of sprint one compares two different things.

Next