Categories Guide

What each capture category looks like in practice — by role. Use this for inspiration when designing your framework.

The Six Categories Are Universal

Everyone gets all six. The question isn't "which categories do I need?" — it's "which categories carry the most weight for my role?" All six catch real things in every professional's life. The difference is which ones you'll use ten times a day vs. once a week.

Category Deep Dives with Real Examples

Questions

What it catches: Anything someone is waiting for an answer on. Unresolved questions are cognitive overhead — they sit in your brain rent-free until someone answers them.

Real examples by role:

Custom fields some roles add:

Classification triggers — how to recognize a Question in the wild:

• Someone says "I need to find out..." or "Can you check on..."
• An email ends with a direct question and no answer yet
• A meeting action item is "get back to [person] about..."
• Someone said they'd "look into it" — that's a question waiting for an answer
• A Slack thread ends with "?" and no resolution

Decisions

What it catches: Choices that were made, who made them, and WHY. This is the category people most regret not tracking. Six months from now, someone will ask "why did we do it this way?" and the decision capture has the answer.

Real examples by role:

Custom fields some roles add:

Classification triggers:

• Someone says "We decided to..." or "Let's go with..."
• A meeting concludes with a chosen direction
• An email thread ends with alignment on an approach
• Someone says "I'm going to..." about a non-trivial choice (not a task — a direction)
• A rejected alternative is mentioned ("We considered X but went with Y")

Tasks

What it catches: Work that has an owner, a deadline, and a definition of done. Everyone tracks tasks. What makes this system different is that tasks are CONNECTED to the other five categories — a task might fulfill a commitment, implement a decision, or answer a question.

Real examples by role:

Custom fields some roles add:

Classification triggers:

• Someone says "I need to..." or "Can you..." followed by concrete work
• An email assigns work with a deadline
• A meeting produces action items
• A commitment implies work ("I'll send that over" = commitment + task)
• A Jira/Asana/Trello ticket is created or assigned

Commitments

What it catches: Promises between people, in BOTH directions. What you promised others AND what others promised you. This is the category most people miss entirely, and it's the one that prevents the most painful failures — broken promises you didn't even know were promises.

Real examples by role:

Custom fields some roles add:

Classification triggers:

• Someone says "I'll...", "I promise...", "I'll get that to you by...", "You'll have it by..."
• Someone says "They said they'd...", "She committed to...", "He's going to..."
• Calendar invites with deliverable expectations
• Email sign-offs with delivery promises ("I'll circle back with numbers on Monday")
• The subtle ones: "Sure, I can do that" in a meeting (this IS a commitment)

The commitment/task split:

When something is both a commitment and a task (common), create both:

• The commitment tracks the promise: who said what to whom, by when
• The task tracks the work: what needs to happen to fulfill the promise

The commitment closes when the promise is fulfilled. The task closes when the work is done. Usually simultaneous, but not always — you might complete the work (task done) but forget to send it (commitment still open).

Context

What it catches: Knowledge with no action attached. Things that aren't tasks or decisions but are important to remember. Context captures are your institutional memory — they make every future decision, classification, and interaction better.

Real examples by role:

Custom fields some roles add:

Classification triggers:

• Information with no action but future value
• "FYI" or "just so you know" messages
• Background in an email that isn't the main point but is worth remembering
• Observations about people's preferences or working styles
• System quirks, historical decisions without formal capture, tribal knowledge
• "The reason we do it this way is..." (context about existing patterns)

Expiration matters:

Context captures should have an Expires field. Some context is permanent ("the client's fiscal year ends in March" — always true). Some expires ("budget freeze until April 15" — stale after that date). Set realistic expirations and let the system clean up stale context.

Projects

What it catches: Named containers that group questions, decisions, tasks, commitments, and context into a coherent body of work. A project has people, a timeline, and a goal.

When to create a project:

When NOT to create a project:

• If it's really just one or two tasks, don't over-organize. Orphan captures are fine.
• If there's no clear end state, it might be a recurring process, not a project.
• If you're not sure, file captures as orphans. You can always move them into a project later when the shape becomes clear.

How Categories Combine By Role

This is the key insight: your role determines which categories carry the most weight, but all six are always present. Here's how typical roles weight them:

The Weights

                    Questions  Decisions  Tasks  Commitments  Context  Projects
Engineer               ●●        ●●●      ●●●       ●●        ●●●       ●
PM                    ●●●         ●●      ●●●      ●●●         ●●      ●●●
Designer               ●●        ●●●      ●●●        ●●       ●●●       ●
Sales                 ●●●          ●       ●●       ●●●       ●●●       ●
Eng Manager            ●●         ●●      ●●●       ●●●       ●●●      ●●
Executive             ●●●        ●●●        ●       ●●●       ●●●      ●●
DevOps/SRE              ●         ●●      ●●●        ●●       ●●●       ●
Freelancer             ●●         ●●      ●●●       ●●●       ●●●      ●●
Product Manager       ●●●        ●●●       ●●        ●●        ●●      ●●●

●●● = primary (core to the role, used daily) ●● = secondary (important but not central) ● = minimal (exists, used occasionally)

Reading the Pattern

• If you make things (engineer, designer, DevOps): Tasks and Context are heavy. You need to know what to build and why things are the way they are.
• If you coordinate things (PM, product manager): Everything is heavy. You're the hub. Commitments and Projects matter most because you're tracking what everyone owes everyone else.
• If you sell things (sales, BD): Commitments and Context are heavy. Relationships run on promises kept and knowledge remembered.
• If you lead people (manager, director): Commitments and Context are heavy. You need to track what you promised your team and what you know about how they work.
• If you decide things (executive, architect): Decisions and Context are heavy. Your most valuable output is well-reasoned choices with documented rationale.

Customizing Your Categories

The six categories are fixed — don't add new top-level categories. If you think you need a seventh category, it's almost certainly a subcategory or a tag within one of the six. Examples:

If after reading this table you STILL think you need a new category — escalate to Josh and Bob. They'll either show you how it maps to the six, or they'll learn something new about the framework that needs to change.

This guide is a reference. Your actual category configuration happens in Phase 2 (Architecture) when the AI designs your system based on your Discovery Document.