Under the Hood

This isn't just ChatGPT
with a Scrum prompt.

Driftless is 35 purpose-built tools grounded in the 2020 Scrum Guide and a 10-dimension maturity framework. Here's how it actually works — and why that matters.

How a coaching session works
01

You open your Jira or Rally board

Driftless lives as a Chrome extension sidebar next to your board. Nothing to set up. No API credentials to configure on day one. It reads the page the same way you do — by looking at what's on screen.

Chrome MV3 Extension · Jira & Rally · Azure DevOps
02

One click reads the entire board

The content script extracts structured ticket data: IDs, titles, assignees, story points, status columns, blockers, labels, and carry-over signals. No data is stored on our servers. The read happens in your browser.

Content Script · Structured DOM Extraction · Zero Server Storage
03

The board data is sent to a purpose-built coaching prompt

This is where Driftless differs from a general AI tool. The data is not sent to a generic "analyze this sprint" prompt. It is routed to one of 35 specific tools — each one written to look for a particular class of signal. Board Maturity Scan looks at 10 coaching dimensions. Team Load Analysis looks at assignee distribution. Sprint Risk looks at carry-overs and blockers older than 5 days. Each prompt is grounded in a specific Scrum Guide accountability or Agile Manifesto principle.

Anthropic Claude · 35 Purpose-Built Prompts · Scrum Guide Grounding
04

The response leads with what to say, not what the data says

Every output follows the same structure: what to do first, why it matters, how to have the conversation, what the evidence is, and which Scrum Guide section it connects to. The framework comes last — because numbers are evidence, not the product. The conversation is the product.

DO → WHY → HOW → EVIDENCE → FRAMEWORK
05

Coaching history builds over time via the Sprint Journal

Every tool response can be saved to the Sprint Journal — a scoped memory that travels with your team profile. Over multiple sprints, patterns become visible. The same blocker type. The same person quietly struggling. The Journal gives Driftless context that makes each subsequent analysis more relevant.

Sprint Journal · Team Profiles · Cross-Sprint Memory
The Maturity Framework

10 dimensions.
Not made up.

The Board Maturity Scan assesses your sprint across 10 dimensions, each scored L0–L4. The dimensions were built from real sprint patterns and mapped back to specific Scrum Guide accountabilities — not invented from theory.

Dimension 01
Story Clarity
Are stories sized, described, and scoped enough to start without a meeting? Unmapped stories are the most common source of mid-sprint confusion.
Dimension 02
Blocker Visibility
Are blockers flagged and being actively worked? Blockers that age past 5 days without a status change are a coaching signal, not a process failure.
Dimension 03
Load Distribution
Is work distributed across the team or concentrated on one or two people? The person carrying everything is a risk to the sprint and to themselves.
Dimension 04
Sprint Goal Alignment
Do the tickets on the board connect to a stated sprint goal? Boards without visible goal alignment drift — story by story, sprint by sprint.
Dimension 05
Carry-Over Rate
What percentage of stories are carry-overs from previous sprints? Recurring carry-overs signal scope problems, not delivery problems.
Dimension 06
WIP Discipline
How many things is each person working on at once? High WIP is the quiet killer of focus and flow that rarely gets named in retros.
Dimension 07
Definition of Done Signals
Do tickets have clear completion criteria? Stories without acceptance criteria create ambiguity that shows up in review, not during the sprint.
Dimension 08
Team Collaboration Patterns
Are stories isolated by assignee or are there visible handoffs and shared ownership? Silo patterns on the board usually reflect silo patterns in the team.
Dimension 09
Estimation Discipline
Are stories estimated consistently? Unestimated or inconsistently estimated stories make sprint planning a guess and velocity a fiction.
Dimension 10
Sprint Health Trajectory
Based on current board state, where is this sprint likely to land? Early warning, not retrospective analysis.
Why it's grounded this way

Every insight traces back to
a real source.

AI coaching tools that make things up are worse than no coaching at all. Driftless grounds every tool output in one of two authoritative sources.

The 2020 Scrum Guide

Every tool in Driftless maps to a specific Scrum Guide accountability — Scrum Master, Product Owner, or Developers. When Driftless flags a blocker pattern, it connects it to the SM's accountability to remove impediments. When it surfaces a sprint goal misalignment, it references the Scrum Team's shared commitment. The framework is visible, not hidden.

The Agile Manifesto

The 12 principles and 4 values of the Agile Manifesto sit underneath every coaching recommendation. "Individuals and interactions over processes and tools" is not a tagline in Driftless — it is the reason the output always leads with a person to talk to, not a metric to fix.

The toolset

35+ tools across
two coaching surfaces.

The extension organizes its tools into two categories that match how coaching actually works: clearing the path so the team can move, and being present in the room when it counts.

Clear the Path
  • Board Maturity Scan (10 dimensions)
  • Sprint Risk Analysis
  • Blocker Identification & Coach
  • Team Load Distribution
  • Carry-Over Pattern Analysis
  • Stakeholder Sprint Summary
  • Velocity Trend Intelligence
  • Tech Debt Signal Detection
  • Definition of Done Review
  • Sprint Goal Alignment Check
Coach the Room
  • Standup Coaching Prompt
  • One-on-One Conversation Starter
  • Retrospective Facilitator
  • Sprint Planning Prep
  • Difficult Conversation Guide
  • Milestone Celebration Framing
  • Team Conflict Early Signal
  • Psychological Safety Check
  • Human Play Follow-Up Loop
  • Sprint Journal Entry & Memory

A note on your data

Driftless reads your board inside your browser. No ticket data, team names, or sprint content is stored on Driftless servers. Board reads happen via your browser's content script and are sent directly to the Anthropic API for analysis. Sprint Journal entries are stored locally in your browser or optionally synced to your own account — never aggregated or analyzed across users.

If your organization has data residency requirements, reach out before starting a trial and we will walk through the architecture with you.

Still have questions?
Ask us directly.

We're in beta and happy to walk through the architecture, the data handling, or anything else with you before you commit to a trial.

Join the Beta → Get in Touch