Elon Musk Skill
First Principles Thinking: solve problems by reasoning from fundamentals, not analogy.
<when_to_use>
When to use
| Situation | Example |
|---|---|
| Only obvious answers appear | "Need revenue growth -> ads?" |
| Cost/structure innovation | "How do we cut infrastructure cost?" |
| Challenge assumptions | "We always do it this way" |
| Strategic decisions | "A vs B decision" |
| Market entry design | "Should we copy competitors?" |
Invocation
/elon-musk SaaS pricing is crowded and undifferentiated
/elon-musk infrastructure cost is 40% of revenue
/elon-musk churn is high and retention playbooks are failing
Outputs
- Assumption deconstruction matrix (constraint vs convention vs unverified)
- Comparison table (analogy-based vs first-principles approach)
- Innovative solution options rebuilt from fundamentals
- Execution plan and risk matrix using inversion/pre-mortem
</when_to_use>
<argument_validation>
Required ARGUMENT check
If $ARGUMENTS is missing, ask immediately:
"Which problem should we deconstruct using first principles?
Examples:
- Cost innovation: 'SaaS infra cost is 40% of revenue'
- Strategy: 'Only generic market-entry options are visible'
- Assumption challenge: 'We always did it this way, but not sure it is right'
- Technical choice: 'Options look similar; need fundamental differences'"
If $ARGUMENTS exists, continue.
</argument_validation>
<core_philosophy>
First Principles Thinking
Core idea
Analogy-based thinking copies and slightly modifies existing patterns. First-principles thinking decomposes to what is fundamentally true, then rebuilds.
| Dimension | Analogy-based | First principles |
|---|---|---|
| Method | Copy and tweak | Decompose then reconstruct |
| Question | "How do others do this?" | "What is undeniably true?" |
| Outcome | Incremental optimization | Potentially disruptive innovation |
| Effort | Lower cognitive load | Higher cognitive load |
| Best fit | Routine decisions | Breakthrough / blocked situations |
Typical cases
| Case | Conventional assumption | First-principles decomposition | Result |
|---|---|---|---|
| Space launch | "Rockets are inherently expensive" | Raw material is a small fraction of sell price | Major cost compression |
| EV battery | "Battery packs are fixed-cost expensive" | Component-level material and process decomposition | New cost curve possibilities |
Supporting methods
| Method | Role | Phase |
|---|---|---|
| Socratic questioning | Assumption deconstruction tool | Phase 2 |
| Inversion | "How do we fail?" reverse risk planning | Phase 4 |
| Pre-mortem | "If we fail in 6 months, why?" | Phase 4 |
</core_philosophy>
<workflow>
Execution flow
| Phase | Task | Tool | Core intent |
|---|---|---|---|
| 0 | Input + environment check | ToolSearch | detect MCP/tool availability |
| 1 | Domain research (convention + facts + innovation cases) | parallel researcher x3 | collect 3 lenses simultaneously |
| 2 | Assumption deconstruction (Socratic questions + A/B/C) | Sequential Thinking (5-7) | separate constraints from habits |
| 3 | Fundamental redesign | parallel analyst x3 | explore 3-5 alternative paths |
| 4 | Execution + risk | Sequential Thinking (3-5) | inversion and pre-mortem |
| 5 | Save + present | Write | .hypercore/first-principles/ |
Complexity profile
| Complexity | Phase 2 depth | Phase 4 depth | Criteria |
|---|---|---|---|
| Simple | 3 thoughts | 2 thoughts | single domain, <=3 assumptions |
| Medium | 5 thoughts | 3 thoughts | multi-domain, 5-7 assumptions |
| Complex | 7+ thoughts | 5 thoughts | industry-level, 10+ assumptions |
<sourcing_strategy>
Search channel strategy
Use Tier-1 MCP channels first, and fallback to built-in web tools when unavailable.
Phase 0: MCP detection
ToolSearch("firecrawl")
ToolSearch("searxng")
ToolSearch("github")
If MCP is available -> pass MCP context into researcher prompts.
If MCP is unavailable -> use WebSearch/WebFetch fallback.
</sourcing_strategy>
<parallel_agent_execution>
Priority: Agent Teams for complex parallel work
- If Agent Teams is available: TeamCreate -> spawn workers -> parallel collaboration.
- If not available: parallel Task calls as fallback.
Phase 1: Parallel domain research (3 directions)
| Agent | Type | Goal | Source requirement |
|---|---|---|---|
| 1-A | researcher | Industry conventions and default approaches | URL + publish date + source type |
| 1-B | researcher | Hard facts/data: cost, benchmarks, constraints | URL + publish date, prefer last 12 months |
| 1-C | researcher | Innovation cases that broke assumptions | URL + publish date |
Phase 3: Parallel alternative-path analysis
| Agent | Type | Evaluation criteria |
|---|---|---|
| 3-A | analyst | feasibility, impact, resources, risk |
| 3-B | analyst | feasibility, impact, resources, risk |
| 3-C | analyst | feasibility, impact, resources, risk |
Model routing
| Phase | Agent | Model | Why |
|---|---|---|---|
| 1 | researcher x3 | sonnet | web-fact collection |
| 2 | Sequential Thinking | main | Socratic decomposition |
| 3 | analyst x3 | sonnet | path feasibility analysis |
| 4 | Sequential Thinking | main | inversion and pre-mortem |
</parallel_agent_execution>
<phase_details>
Phase 2: Assumption deconstruction (core)
Socratic 6-question set
Apply to each convention collected in Phase 1:
| Question type | Prompt | Purpose |
|---|---|---|
| Clarification | "What exactly does this mean?" | make assumption explicit |
| Assumption probe | "What are we assuming?" | reveal hidden premises |
| Evidence test | "What evidence proves this?" | separate fact from belief |
| Perspective shift | "What other frame exists?" | break fixation |
| Implication test | "If false, what changes?" | evaluate removal impact |
| Meta question | "Why did we accept this as normal?" | expose inertia |
Required A/B/C classification
| Class | Mark | Definition | Treatment |
|---|---|---|---|
| Physical/technical constraints | A | natural law, hard technical limits | keep as fundamentals |
| Habitual convention | B | "because industry does it" | remove |
| Needs validation | C | testable but not yet confirmed | verify with search/data |
Phase 2 sequential-thinking template
thought 1: extract 5-10 assumptions from Phase 1 material
thought 2: apply Socratic questions and initial A/B/C labels
thought 3: cross-check C items with known facts
thought 4: run extra search for unresolved C items
thought 5: finalize classification and deconstruction matrix
Phase 3: Fundamental redesign
Input
- A items: true constraints
- B items: removed conventions
Process
1. Keep only A constraints; strip B assumptions.
2. Re-ask: "If we design from these truths only, what changes?"
3. Explore 3-5 alternatives in parallel.
4. Build "current vs first-principles" comparison table.
5. Select best path with explicit rationale.
Phase 4: Execution and risk
| Method | Question | Process |
|---|---|---|
| Inversion | "How do we fail for sure?" | enumerate 5-7 failure paths and map preventions |
| Pre-mortem | "If we fail in 6 months, why?" | identify top causes and preemptive controls |
</phase_details>
<document_storage>
Result storage
| Item | Rule |
|---|---|
| Path | .hypercore/first-principles/[number].[problem_summary].md |
| Numbering | `ls .hypercore/first-principles/ |
| Flow | ensure folder -> Write -> return saved path |
</document_storage>
<result_structure>
Output structure
| Section | Content |
|---|---|
| Header | date, problem statement |
| 1. Domain snapshot | conventions, facts, innovation cases (URLs required) |
| 2. Deconstruction matrix | assumptions + Socratic checks + A/B/C + evidence |
| 3. Fundamental redesign | current vs first-principles comparison and chosen path |
| 4. Execution + risk | action plan, inversion failures, pre-mortem |
| Sources | full URL list |
</result_structure>
<examples>
Core example: cost innovation
User: /elon-musk infrastructure cost is 40% of SaaS revenue
Phase 0: detect MCP/tools
Phase 1: researcher x3 in parallel
- conventions: "standard cloud cost structure"
- facts: compute/network unit economics (recent data)
- innovation cases: cost-disruptive infra companies
Phase 2: Sequential Thinking
assumption 1: "Current provider is always optimal" -> B
assumption 2: "Serverless is always cheaper" -> C -> B after verification
assumption 3: "Latency constraints are physical" -> A
Phase 3: analyst x3 parallel alternatives
A: bare metal + automation
B: hybrid reserved + serverless cold path
C: edge compute migration
Phase 4: inversion + pre-mortem
risk: migration downtime -> blue/green rollout
risk: ops complexity -> automation-first controls
Save: .hypercore/first-principles/00.saas_infra_cost_innovation.md
<validation>
Validation checklist
| Item | Required |
|---|---|
| ARGUMENT | Ask if missing |
| Phase 0 | MCP/tool availability checked |
| Phase 1 | researcher x3 (convention + facts + cases), source grading |
| Phase 2 | Sequential Thinking 3+ steps, 5+ A/B/C items, 2+ fact cross-checks |
| Phase 3 | analyst x3, comparison table included |
| Phase 4 | inversion 3+ risks, pre-mortem included |
| Save | .hypercore/first-principles/ |
| Sources | URL + date + source type for factual claims |
| Forbidden |
|---|
| Start without ARGUMENT |
| Deconstruct assumptions without domain research |
| Redesign without A/B/C matrix |
| Execution plan without inversion/pre-mortem |
| Factual claims without sources |
| Pure analogy/copy strategy |
| Exit without saving results |