Bootstrap Learning.instructions

---
type: instruction
lifecycle: stable
inheritance: inheritable
description: "Domain-agnostic learning and rapid context acquisition — from zero to partnership fast"
application: "When entering a new project, domain, or user relationship"
applyTo: "**/*"
currency: 2026-04-27
---

# Bootstrap Learning

Rapid context acquisition protocol. Go from zero knowledge to effective partnership fast.

## The Bootstrap Problem

Learning a new domain/project is hard because you don't know what you don't know. This instruction provides systematic orientation.

## First Session Protocol

When entering a new project:

### 1. Orient (30 seconds)

| Check | What to scan |
|-------|-------------|
| README.md | Project purpose, setup, constraints |
| package.json / pyproject.toml | Tech stack, dependencies, scripts |
| Directory structure | Architecture pattern |
| Recent commits | Active work areas |

### 2. Understand the Goal

Ask yourself (or the user):
- What is this project?
- What problem does it solve?
- Who uses it?
- What's the current state?

### 3. Learn the User

| Signal | Inference |
|--------|-----------|
| Terse messages | Match brevity, skip preamble |
| Detailed questions | Provide depth, explain reasoning |
| Technical vocabulary | Use technical terms freely |
| Plain language | Explain jargon when used |
| "Just do X" | Action-oriented, minimize discussion |
| "What do you think?" | Values collaboration, share perspective |

### 4. Engage

Only after orientation: propose, build, help.

## Learning Methodology — 5 Phases

### Phase 1: Discovery — Map the territory

- Boundary mapping: "What does X include and exclude?"
- Vocabulary scan: "What are the 5 key terms?"
- Adjacent domains: "What's related but different?"

**Exit criteria**: Can describe the domain in one sentence.

### Phase 2: Foundation — Nail the core concepts

- Ask for simplest explanation of each core concept
- Demand concrete examples, not abstractions
- Test by explaining back in own words
- **Red flag**: If explanation uses domain jargon, haven't bottomed out

**Exit criteria**: Can explain core concepts without jargon.

### Phase 3: Elaboration — Add depth through cases

| Type | Purpose |
|------|---------|
| Happy path | How it works normally |
| Edge cases | Where it breaks |
| Anti-patterns | Common mistakes |
| Trade-offs | When NOT to use it |

**Exit criteria**: Can identify when to use and when NOT to use.

### Phase 4: Connection — Link to existing knowledge

- Map analogies: "This is like [X] because..."
- Find contradictions: "This conflicts with [Y] — which is right?"
- Identify synergies: "Combining with [Z] could improve..."

**Exit criteria**: At least 2 connections to known concepts.

### Phase 5: Consolidation — Remember what matters

- Note key learnings in session memory
- If persistent memory available, write durable notes
- Flag gaps that remain

## Gap Identification

| Signal | Type of Gap | Action |
|--------|-------------|--------|
| "I don't know the right question" | Vocabulary gap | Return to Phase 1 |
| "I understand words but not concept" | Foundation gap | Return to Phase 2 |
| "I understand but can't apply" | Elaboration gap | Return to Phase 3 |
| "I know this but it feels isolated" | Connection gap | Phase 4 |
| "I keep re-learning this" | Consolidation gap | Phase 5 |

## The Feynman Check

> If you can't explain it simply, you don't understand it well enough.

After learning a concept, try to explain it in one paragraph using no jargon. If you can't, identify which part is unclear and loop back.