Team Analysis
Team Analysis connects to your Jira or Azure DevOps board and produces a comprehensive analysis of your team's real delivery patterns. The analysis becomes a reusable profile that calibrates future planning sessions.
What gets analyzed
| Category | Details |
|---|---|
| Velocity | Per-sprint velocity, team average, per-developer breakdown with delivery vs KTLO split |
| Sprint Cadence | Sprint length, seasonal patterns, velocity trends over time |
| Story Structure | Average stories per epic, story point distribution, typical story size |
| Task Patterns | Average tasks per story, type distribution (Development %, Testing %, Deploy %) |
| Acceptance Criteria | AC count per story, writing style (Given/When/Then vs flexible), coverage patterns |
| Definition of Done | Team's actual DoD items extracted from completed stories |
| Naming Conventions | Epic naming style (quarter-scoped, feature-based), ticket organization patterns |
| Team Members | Per-contributor analysis with velocity, discipline (backend/frontend/etc.), stories completed, sprint participation |
| Estimation Bias | Whether the team tends to over- or under-estimate |
| Board Workflow | Column workflow analysis showing how work items flow through states |
How it works
- Select a board — pick your Jira project or Azure DevOps team board
- Choose sprint scope — analyze last 3, 5, 10, or all sprints
- Review the analysis — 6-page walkthrough: Instructions, Sample Epic, Sample Stories, Sample Tasks, Sample Sprint Plan, Analysis Report
- Save as a profile — stored in SQLite, reusable across planning sessions
Per-developer breakdown
Each team member gets individual analysis:
| Metric | Description |
|---|---|
| Velocity | Individual points per sprint average |
| Discipline | Primary skill area (backend, frontend, fullstack, infrastructure, etc.) — LLM-classified from actual work |
| Stories completed | Total stories delivered in the analysis window |
| Sprint participation | Number of sprints active |
| Delivery vs KTLO | Split between feature delivery and keeping-the-lights-on work |
Analysis exports
Analysis results can be exported as HTML or Markdown reports. Both formats include all analysis sections plus a Team Members table. When an analysis profile is used during planning, the export includes a provenance note linking back to the analysis.
Analysis-Calibrated Planning
When you start a planning session, you can select a saved analysis profile. This calibrates the entire planning pipeline to match your team's real patterns.
Intake auto-fill
The analysis profile auto-fills intake questions so you don't have to answer them manually:
| Question | Source |
|---|---|
| Q6 — Team size | Replaced with a team member multi-select picker showing each developer's velocity and discipline |
| Q7 — Team roles | Auto-filled from selected members' disciplines |
| Q8 — Sprint length | Derived from sprint date ranges in the analysis |
| Q9 — Velocity | Calculated from selected team members' individual per-sprint averages |
| Q11 — Tech stack | Auto-filled from the analysis profile's tech stack (shown as suggestion, user can override) |
| Q12 — Integrations | Auto-filled from the analysis profile's integrations list |
| Q27 — Sprint selection | Falls back to analysis sprint data when live tracker is unavailable |
Team member multi-select
When an analysis profile has contributor data, Q6 (team size) becomes a multi-select picker instead of a free-text number:
- Each member shows their velocity (pts/sprint) and discipline
- Select specific developers for this project with Space, confirm with Enter
- Velocity is calculated from the selected members' individual averages
- Team roles (Q7) auto-populated from selected members' disciplines
Calibration banners
Each pipeline stage shows a calibration banner explaining what analysis data influenced the output:
| Stage | Banner shows |
|---|---|
| Project Analyzer | Selected profile name and source board |
| Story Writer | Target stories/epic, AC count, points scale, DoD source |
| Task Decomposer | Target tasks/story, type distribution (Dev/Test/Deploy %) |
| Sprint Planner | Selected team members with individual velocities, total team velocity |
What gets calibrated
| Artifact | Calibration |
|---|---|
| Epics | LLM-reformatted to match team's naming convention (quarter-scoped, feature-based, etc.) and template sections |
| Stories per feature | Matches team's historical average stories per epic |
| AC count | Enforces team's median acceptance criteria count per story |
| AC format | Respects team's writing style (Given/When/Then or flexible) |
| Definition of Done | Uses team's actual DoD items instead of generic checklist |
| Story points | Includes confidence scoring and references to similar stories from the analysis |
| Tasks per story | Matches team's average task count |
| Task type distribution | Matches team's Development/Testing/Deploy percentages |
Epic review
Before the feature generator runs, an epic review page lets you review and edit the project epic. When an analysis profile is active, the epic is LLM-reformatted to match your team's style:
- Quarter-scoped naming when the team uses that convention (e.g., "Q2 2026 — Customer Portal")
- Template sections matching the team's epic structure
- Sync to Jira or Azure DevOps directly from the review page