Elon Hardcore

Use when reviewing product, strategy, hiring, execution, operations, prioritization, org design, manufacturing/process, or messaging decisions through an Elo...

Security Scan

Instruction-only review skill whose requests and instructions align with its stated purpose and that does not ask for credentials, install software, or access unrelated system resources.

Purpose & Capability

The name/description match the SKILL.md and framework: the skill is a first-principles decision-review checklist. It is instruction-only and asks for no binaries, env vars, or external services — proportional to the described purpose.

Instruction Scope

Runtime instructions are limited to: identify a decision object, apply the checklist in references/framework.md, and produce one of several specified output formats. The skill instructs the agent to read only the included references/framework.md file for full reviews. It does not instruct the agent to read unrelated files, call external endpoints, or access system credentials.

Install Mechanism

No install spec or code is present (instruction-only). Nothing is downloaded or written to disk by the skill itself.

Credentials

The skill declares no required environment variables, credentials, or config paths. There are no requests for secrets or unrelated credentials.

Persistence & Privilege

Flags: always is false and model invocation is allowed (platform default). The skill does not request permanent presence or elevated system privileges and does not modify other skills' settings.

Assessment

This skill is internally coherent and low technical risk because it's instruction-only and asks for no credentials or installs. Main non-technical risks are from the outputs it generates: the guidance is intentionally blunt and can recommend deleting processes, firing people, or making aggressive decisions — do not treat its verdicts as legal, HR, or compliance advice without human oversight. Practical steps before using: 1) Test with non-sensitive examples to verify tone and style. 2) Avoid pasting secrets or private credentials into prompts you ask the skill to review. 3) If you will act on hiring/termination or regulatory decisions, route outputs to legal/HR for review. 4) If you want guardrails, add prompt constraints (e.g., require risk/impact sections, legal/ethical checks, or a softer tone). 5) Monitor agent outputs when the agent can invoke the skill autonomously to ensure it doesn't produce harmful operational actions based solely on the skill's blunt recommendations.

Elon Review

Use this skill to pressure-test decisions for any agent in a blunt, high-signal, first-principles style.

Core workflow

1. Identify the decision object: product idea, strategy, plan, org/process, hiring call, message, or prioritization choice.
2. Reduce the problem to fundamentals. Separate hard constraints (physics, math, unit economics, actual user behavior, time) from inherited assumptions.
3. Run the review against the checklist in references/framework.md.
4. Flag analogy-thinking, bureaucracy, and optimization of things that should not exist.
5. Return a concise verdict with required actions, not vague commentary.

Output modes

Pick the lightest mode that fits the request.

1) Quick verdict

Use for short asks or chat replies.

Output:

• Verdict: green / yellow / red
• One-sentence assessment
• Top 3 issues
• Next move

2) Full review

Use for product, strategy, operations, hiring, or resource allocation decisions.

Output:

• Status + confidence
• Assessment
• What is true at first principles
• Principle violations
• Bottlenecks / anti-patterns
• Required actions
• What to cut / simplify / accelerate

3) First-principles rewrite

Use when the input is fuzzy, overcomplicated, or too marketing-heavy.

Output:

• Original flaw summary
• Reframed problem
• Cleaner first-principles version
• What changed

4) Brutal questions

Use when the user wants challenge, not answers.

Output 5-15 hard questions that expose weak assumptions, fake constraints, missing speed, bad incentives, or lack of real usefulness.

Review rules

• Prefer blunt clarity over soft hedging.
• Do not mimic Elon’s personality; mimic the decision rigor.
• Do not treat regulations, org charts, incumbent habits, or industry norms as sacred.
• Distinguish between:
• immutable constraints
• temporary constraints
• fake constraints
• If something should be deleted, say so before suggesting optimization.
• If a plan is too slow, say so explicitly.
• If the proposal is solving a trivial problem, say that too.

Domain-specific emphasis

Product / strategy

Focus on usefulness, magnitude of impact, hidden assumptions, rate of iteration, and whether the problem matters.

Operations / execution

Focus on bottlenecks, cycle time, serial vs parallel work, communication drag, and whether teams are optimizing around reality or process theater.

Hiring

Focus on proof of exceptional ability, evidence of solving hard problems, depth of understanding, and attitude toward difficulty.

Messaging / writing

Focus on signal density, truthfulness, whether the message hides weak thinking, and whether it states the real problem and real value.

Mandatory reference

Read references/framework.md when doing a full review or first-principles rewrite.

Default response template

[STATUS: GREEN / YELLOW / RED]
[CONFIDENCE: 0.00-1.00]

ASSESSMENT:
<1-2 sentence blunt judgment>

FIRST-PRINCIPLES TRUTH:
- <what is actually true>

PRINCIPLE VIOLATIONS / RISKS:
- <violation or anti-pattern>

BOTTLENECK:
- <primary bottleneck>

REQUIRED ACTIONS:
1. <direct action>
2. <direct action>
3. <direct action>

WHAT TO CUT / SIMPLIFY / ACCELERATE:
- Cut:
- Simplify:
- Accelerate: