First Principles Analysis
Description
Break complex problems down to fundamental truths, then rebuild solutions from the ground up.
Trigger
User asks for a decision, strategy advice, "how should I approach", "what's the best way to", or any complex problem with multiple variables.
Process
- Identify constraints — Ask: What's fixed? Budget, timeline, team size, tech stack, regulations. List them explicitly.
- Find root cause — Strip away assumptions. Ask "why" until you hit bedrock. Most problems are symptoms of deeper issues.
- Generate 3 solutions from different angles:
- The fast/cheap path (minimize resources)
- The thorough/expensive path (maximize quality)
- The lateral path (reframe the problem entirely)
- Attack each solution — For every option, name the top 2 risks and the conditions under which it fails.
- Check memory for precedent:
secondBrain.search for similar past decisions- Surface what worked/failed before
- Recommend one path — State the recommendation, the key assumption it depends on, and the trigger to reconsider.
Output
- Constraints list (3-5 bullets)
- Root cause (1 sentence)
- 3 options table (option / upside / risk / timeline)
- Recommendation with confidence level (high/medium/low)
- Revisit trigger: "Reconsider if X changes"
Examples
- "Should I hire a contractor or build in-house?" — Constraints: budget $5k/mo, need it in 2 weeks, core product feature. Root cause: speed vs. ownership tradeoff. Recommend contractor with handoff plan.
- "How do I fix my funnel conversion?" — Constraints: 2% conversion, $50 CAC, landing page + email sequence. Root cause: messaging mismatch (traffic intent vs. page promise). Three angles: copy rewrite, audience retargeting, offer restructure.
Failure Modes
- Too abstract — If the user gives a vague problem, ask for specifics before running the framework. Don't first-principles a grocery list.
- Analysis paralysis — Cap at 3 options. More than that means the framing is wrong.
- Missing context — Always check
secondBrain.search first. The user may have already decided this and forgotten.
- Overconfidence — If constraints are uncertain, say so. "Medium confidence — this depends on X being true."
1---2name: first-principles3description: First Principles Analysis4---5# First Principles Analysis67## Description8Break complex problems down to fundamental truths, then rebuild solutions from the ground up.910## Trigger11User asks for a decision, strategy advice, "how should I approach", "what's the best way to", or any complex problem with multiple variables.1213## Process141. **Identify constraints** — Ask: What's fixed? Budget, timeline, team size, tech stack, regulations. List them explicitly.152. **Find root cause** — Strip away assumptions. Ask "why" until you hit bedrock. Most problems are symptoms of deeper issues.163. **Generate 3 solutions from different angles:**17 - The fast/cheap path (minimize resources)18 - The thorough/expensive path (maximize quality)19 - The lateral path (reframe the problem entirely)204. **Attack each solution** — For every option, name the top 2 risks and the conditions under which it fails.215. **Check memory for precedent:**22 - `secondBrain.search` for similar past decisions23 - Surface what worked/failed before246. **Recommend one path** — State the recommendation, the key assumption it depends on, and the trigger to reconsider.2526## Output27- Constraints list (3-5 bullets)28- Root cause (1 sentence)29- 3 options table (option / upside / risk / timeline)30- Recommendation with confidence level (high/medium/low)31- Revisit trigger: "Reconsider if X changes"3233## Examples341. "Should I hire a contractor or build in-house?" — Constraints: budget $5k/mo, need it in 2 weeks, core product feature. Root cause: speed vs. ownership tradeoff. Recommend contractor with handoff plan.352. "How do I fix my funnel conversion?" — Constraints: 2% conversion, $50 CAC, landing page + email sequence. Root cause: messaging mismatch (traffic intent vs. page promise). Three angles: copy rewrite, audience retargeting, offer restructure.3637## Failure Modes38- **Too abstract** — If the user gives a vague problem, ask for specifics before running the framework. Don't first-principles a grocery list.39- **Analysis paralysis** — Cap at 3 options. More than that means the framing is wrong.40- **Missing context** — Always check `secondBrain.search` first. The user may have already decided this and forgotten.41- **Overconfidence** — If constraints are uncertain, say so. "Medium confidence — this depends on X being true."
GodMode-Team