samarv/people-profile-builder

People Profile Builder

Overview

Create or update a per-person profile to help tailor communication and collaboration. Profiles are stored as Markdown files in input/org/profiles/ using a consistent template.

Workflow

1) Identify the person

• Use input/org/colleagues.json to confirm the canonical name and any aliases.
• If the name is ambiguous or missing, ask one question at a time until clear.
• Do not invent details; only use user-provided information.

2) Determine the file path

Store each profile at:

input/org/profiles/<slug>.md

Slug rules:

• Lowercase
• Spaces to hyphens
• Remove diacritics (for ascii-only filenames)
• Example: Hans Magnus Inderberg -> hans-magnus-inderberg.md

3) Create or update the profile

• If the file does not exist, create it using the template below.
• If it exists, update only the relevant sections and preserve existing content.
• If a section has no information yet, leave a - TBD placeholder.

4) Evolve patterns toward principles

As you gather more evidence:

1. Flag "why" moments: When someone explains their reasoning, tag with [WHY STATED]
2. Look for consistency: Do they state similar reasoning in different contexts?
3. Promote to principle: After 3+ confirming instances, create entry in "Stated Operating Principles"
4. Track predictions: Note when principles correctly/incorrectly predict behavior
5. Maintain honesty: Only claim "principle" status when you have their explicit reasoning, not your inference

Quality gates for promoting to principle:

• [ ] Do I have their words explaining "why"? (Not just my inference)
• [ ] Does this appear in 3+ different contexts?
• [ ] Can I use this to predict future behavior?
• [ ] Would they recognize this if I showed them?

If all YES → It's a principle worth documenting

Profile Template

# <Full Name>

## Overview
- Role:
- Team:
- Relationship:

## Communication Style
-

## Tone Preferences
-

## Decision Style
-

## Motivations
-

## Sensitivities / Watch-outs
-

## Preferred Channels
-

## Response Cadence
-

## Working With Them
-

## Stated Operating Principles

*Principles the person has explicitly articulated. Only add when you have 3+ instances confirming consistency.*

### [Principle Name]

**Statement**: [One-sentence principle in their words or close paraphrase]

**Evidence**:
- Quote: "[Direct quote]" (Source, Date)
- Quote: "[Another instance]" (Source, Date)
- Context: [Related background]

**How this shows up**:
- [Behavioral manifestation 1]
- [Behavioral manifestation 2]
- [Behavioral manifestation 3]

**Predictive value**:
- [What this principle helps you anticipate]
- [Situations where this principle will drive their behavior]

**Verified predictions**:
- ✓ [Prediction that came true] (Date)
- ✗ [Prediction that failed] (Date) - *[What this taught you]*

**Confidence**: High/Medium/Low

## Principle Tensions

*Optional: Add only when you've observed 2+ principles that conflict*

### [Tension Name]: [Principle A] vs [Principle B]

**Principle A**: "[Statement]"
- Evidence: [Quote/behavior]

**Principle B**: "[Statement]"
- Evidence: [Quote/behavior]

**When they conflict**:
- [Describe the situations where both principles apply]

**Resolution pattern**: [How they typically resolve this tension]
- Example: [Specific instance]

## Good Communication Examples
-

## Bad Communication Examples
-

## Notes / Examples

*Tag special entries to track principle evolution:*
- `[WHY STATED]` - Person explicitly explains reasoning
- `[PRINCIPLE CONFIRMED]` - Existing principle appears again
- `[PRINCIPLE CONFLICT]` - Behavior contradicts stated principle
- `[OPEN QUESTION]` - Unexplained behavior worth investigating

---

Profile Evolution Over Time

Monthly Review Process

Profiles should evolve from behavioral patterns to predictive mental models over time.

1. Search for principle evidence:

   grep -r "\[WHY STATED\]" input/org/profiles/
   

2. Cluster by theme: Group related "why" statements across different interactions

3. Promote when ready: Create "Stated Operating Principles" entry when:

   • 3+ instances confirming consistency
   • Applies across multiple contexts
   • Enables behavioral prediction

4. Validate predictions: Track when principle-based predictions succeed/fail

Confidence Calibration

• High: Person explicitly stated this, appears in 3+ contexts, predictions validated
• Medium: Person stated this once or twice, pattern consistent, predictions untested
• Low: Inferred from behavior, not explicitly stated, needs verification

When You Don't Know the "Why"

For unexplained behaviors, add to Notes:

- [OPEN QUESTION] Why does [Person] [behavior]?
  - *Possible reasons*: [Hypothesis 1], [Hypothesis 2]
  - *How it affects working with them*: [Observable impact]
  - *How to test*: [Question to ask or situation to observe]

This maintains intellectual honesty: you acknowledge what you don't know rather than fabricating explanations.

Example: Pattern → Principle Evolution

Month 1: Behavioral Pattern

## Communication Style
- Values early sharing of drafts for feedback before polish

Month 2: "Why" Captured

## Notes / Examples
- [WHY STATED] "Would love your opinions when reading through it so I can make it do a better pass, with all our input" (Feb 6 2026)

Month 3: Multiple Instances

## Notes / Examples
- [WHY STATED] "Would love your opinions..." (Feb 6 2026)
- [PRINCIPLE CONFIRMED] Shared Scenario vision before reading it himself (Feb 6 2026)
- [PRINCIPLE CONFIRMED] Iterated quickly based on Carl's feedback (Feb 7 2026)

Month 4: Promoted to Principle

## Stated Operating Principles

### Collaborative Refinement

**Statement**: "Better outcomes come from collective input on early drafts"

**Evidence**:
- Quote: "Would love your opinions when reading through it so I can make it do a better pass, with all our input" (Slack, Feb 6 2026)
- Shares documents in draft form, not polished final (Scenario vision doc, Feb 2026)
- Iterates quickly based on feedback: "Doing another iteration now" (Feb 7 2026)

**How this shows up**:
- Shares work-in-progress openly
- Explicitly invites critique before polish
- Commits to fast iteration cycles

**Predictive value**:
- Will share documents in draft form, not polished final
- Expects collaborative iteration, not one-way review
- Values speed of feedback over perfect first draft

**Verified predictions**:
- ✓ Shared Scenario doc before self-review (Feb 2026)
- ✓ Iterated same day based on Carl's feedback (Feb 2026)

**Confidence**: High