first-principles
/first-principles
Clarify a problem before planning through a brief, source-backed interview. Use when the user asks for first-principles thinking, wants to challenge assumptions, or wants to establish what actually needs solving before doing work. Produce an agreed problem frame and one next step, not an implementation plan.
Assessment and security
- Files in this package
- 1
- Scripts
- 0
- HTML comments
- 0
- URLs in SKILL.md
- 0
No signals detected. Repository changed on 25.10.2026, without references to retired models or APIs.
MIT, open: use and modify while retaining author attribution.
What it does and when to use it
Clarify a problem before planning through a brief, source-backed interview. Use when the user asks for first-principles thinking, wants to challenge assumptions, or wants to establish what actually needs solving before doing work. Produce an agreed problem frame and one next step, not an implementation plan.
Original author description
Clarify a problem before planning through a brief, source-backed interview. Use when the user asks for first-principles thinking, wants to challenge assumptions, or wants to establish what actually needs solving before doing work. Produce an agreed problem frame and one next step, not an implementation plan.
How it works
- Interview
- Finish
SKILL.md preview
--- name: first-principles version: 1.0.0 description: Clarify a problem before planning through a brief, source-backed interview. Use when the user asks for first-principles thinking, wants to challenge assumptions, or wants to establish what actually needs solving before doing work. Produce an agreed problem frame and one next step, not an implementation plan. --- # First Principles Establish the minimum justified premises needed to proceed. Challenge the frame, not every design decision. ## Interview - Read existing context first. Separate the proposed solution from the beneficiary, desired outcome, and observable success. Do not invent targets. - Actively check factual premises against sources: connected documents, records, or code for internal claims; current primary web sources for external claims. Retrieve before asking questions that sources can answer. Keep confidential details out of public searches. - Cite evidence beside each factual claim. Never use model memory as verification. Check relevance and freshness; expose conflicting evidence. Label unsupported claims as user-reported, assumptions, or unknowns. Treat goals and preferences as user choices, not externally verifiable facts. Label deductions as inferences. - Ask one short, high-leverage question per turn, with at most one sentence of context. Ask only when the answer could change the outcome, a real constraint, or the next step. Do not repeat answered questions or ask the user to do retrievable research. - Test the most consequential assumption: what supports it, what would disprove it, and what changes without it? Distinguish hard constraints from preferences and inherited conventions. Offer tentative interpretations for correction without steering the user toward agreement. - Update the frame after each answer. Discard conclusions dependent on a corrected premise. Keep research focused on decision-changing claims; stop searching once adequately supported. When evidence is unavailable, state the gap rather than filling it. ## Finish Stop when the outcome, constraints, and consequential uncertainty are explicit, not when every design choice is settled. Keep unresolved items visible; uncertainty may make validation the next step. Return about 100-150 words, excluding citations, using these labels: - **Outcome:** Who needs what change, and what would count as success? - **Facts:** Only source-supported claims; cite them, or state none verified. - **Constraints:** Genuine limits and their basis; distinguish preferences. - **Assumptions / unknowns:** Unsupported premises and decision-changing gaps. - **Implication + next step:** What follows from the premises, conditional where needed; exactly one proposed action or test. Ask for one correction or confirmation of the frame before handing off. Never equate agreement with verification. Do not produce an implementation plan, execute the next step, or modify project files.
Versions
| Repository | Score | Licence |
|---|---|---|
| This page · ccomkhj/skills | 90 | MIT |
| aleknitka/agent-creative-thinking-skills | 100 | MIT |
| swap-cameroon/kuro-antigravity-config | 100 | MIT |
| m1nga/first-principles | 100 | MIT |
| coffeesam/claude-futureman | 100 | MIT |
| masihmoafi/skills | 100 | MIT |
| v60samurai/first-principles | 100 | MIT |
| deciqAI/knowledge-skills | 100 | MIT |
| ajbmachon/ajbm-skills | 100 | MIT |
Database snapshot: 13.11.2026 · package version 74ad76d6bb17 (SHA-256 digest of the entire package)
Downloads in the last 30 days: 0 · ranking
Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.
Similar skills
- 100first-principles-thinking
Systematic first principles thinking for any problem domain. Use when the user says "analyze from first principles", "第一性原理", "从根本分析", "从零开始思考", "think from scratch", "question this design", "is this the right approach", "challenge assumptions", "挑战假设", "为什么要这样做", "有没有更好的方案",… - 100first-principles-thinking
Systematic first principles thinking for any problem domain. Use when the user says "analyze from first principles", "第一性原理", "从根本分析", "从零开始思考", "think from scratch", "question this design", "is this the right approach", "challenge assumptions", "挑战假设", "为什么要这样做", "有没有更好的方案",… - 100problem-definition
Frame the real user or business problem (not a code/test bug) using Jobs-To-Be-Done, the 5-whys ladder, and problem-vs-symptom separation. Use when the user asks "are we solving the right problem", wants to write a problem statement, needs to separate problems from symptoms,… - 100first-principles-thinking
Use when assumptions, analogies, legacy constraints, or accepted ways of working may be hiding a better solution. Use when applying First Principles Thinking, reasoning from fundamentals, assumption stripping, rebuilding from basic truths, or non-incremental problem solving. - 90bmad-cis-problem-solving
Apply systematic problem-solving methodologies to complex challenges. Use when the user says "guide me through structured problem solving" or "I want to crack this challenge with guided problem solving techniques - 90bmad-cis-problem-solving
Apply systematic problem-solving methodologies to complex challenges. Use when the user says "guide me through structured problem solving" or "I want to crack this challenge with guided problem solving techniques