メニュー
Applies hypothesis-test-analyze loops to narrow uncertainty with evidence. Use when a problem is iterative, multiple hypotheses are plausible, or experimentation is the right way to learn.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Stress-tests an idea through adversarial analysis, steelmanning, and concrete counterarguments. Use when claims need pressure-testing, failure modes may be subtle, or security and robustness matter.
Pressure-tests a proposed design through research, architecture, engineering, and cold-reader review before implementation. Use when a non-trivial plan exists and you want to find weak assumptions before writing code.
Runs the hypothesis → benchmark → keep/discard protocol for an autoresearch iteration. Use when executing a `shaka autoresearch` loop; it is loaded automatically on each iteration.
Evaluates an external source (repo, tool, article, talk) against the current project and produces a report-only steal/adapt/reject decision table with named rejections. Use when mining another system for ideas worth adopting without copying its scope.
Creates the benchmark harness for a new autoresearch run from a natural-language objective. Use when starting an interactive `shaka autoresearch start` setup session.
Expands the solution space with structured creative exploration and divergent thinking. Use when the task benefits from novel options, brainstorming, or deeper exploration beyond the first obvious answer.
| name | science |
| description | Applies hypothesis-test-analyze loops to narrow uncertainty with evidence. Use when a problem is iterative, multiple hypotheses are plausible, or experimentation is the right way to learn. |
| key | science |
| include_when | Iterative problem. Experimentation needed. Multiple hypotheses to test. |
Apply the scientific method with hypothesis formation, controlled testing, and analysis of results. Ideal for problems that require systematic experimentation.
Route to the appropriate workflow based on the request.
| Trigger | Workflow |
|---|---|
| Full scientific investigation | Workflows/Investigate.md |
| Quick hypothesis test | Workflows/QuickTest.md |
| Workflow | Purpose | Output |
|---|---|---|
| Investigate | Full scientific method cycle | Hypothesis → Experiment → Analysis → Conclusion |
| QuickTest | Rapid hypothesis validation | Single hypothesis → Test → Result |
┌─────────────────────────────────────────────────────────┐
│ STEP 1: OBSERVE │
│ Gather data about the current state │
│ What do we know? What patterns exist? │
└─────────────────────────────────────────────────────────┘
↓
┌─────────────────────────────────────────────────────────┐
│ STEP 2: HYPOTHESIZE │
│ Form testable predictions │
│ "If X, then Y because Z" │
└─────────────────────────────────────────────────────────┘
↓
┌─────────────────────────────────────────────────────────┐
│ STEP 3: DESIGN │
│ Create experiments with controls │
│ What will you measure? What's the control? │
└─────────────────────────────────────────────────────────┘
↓
┌─────────────────────────────────────────────────────────┐
│ STEP 4: EXECUTE │
│ Run experiments systematically │
│ Document everything, avoid bias │
└─────────────────────────────────────────────────────────┘
↓
┌─────────────────────────────────────────────────────────┐
│ STEP 5: ANALYZE │
│ Interpret results objectively │
│ Did results support or refute hypothesis? │
└─────────────────────────────────────────────────────────┘
↓
┌─────────────────────────────────────────────────────────┐
│ STEP 6: ITERATE │
│ Refine hypotheses based on findings │
│ New questions → New cycle │
└─────────────────────────────────────────────────────────┘
## Scientific Analysis: [Topic]
### Observation
- **Current State**: [What we know]
- **Patterns**: [What we notice]
- **Anomalies**: [What's unexpected]
### Hypotheses
| # | Hypothesis | Testable Prediction | Priority |
| --- | ------------- | -------------------- | ------------ |
| 1 | [If X then Y] | [Measurable outcome] | High/Med/Low |
### Experiment Design
- **Variable**: [What we're changing]
- **Control**: [What stays constant]
- **Measurement**: [How we'll know]
### Results
| Hypothesis | Prediction | Actual | Supported? |
| ---------- | ---------- | ---------- | -------------- |
| H1 | [Expected] | [Observed] | Yes/No/Partial |
### Conclusion
- **Finding**: [What we learned]
- **Confidence**: [How certain]
- **Next Steps**: [Follow-up experiments or actions]
Works well with:
Example 1: Performance debugging
"Why is the API slow on Mondays?"
→ Observe: Response times spike 3x on Monday mornings
→ Hypothesize: H1: Cache cold after weekend. H2: Traffic spike. H3: Batch jobs.
→ Test: Monitor cache hits, traffic volume, job scheduler
→ Result: H1 supported - cache hit rate drops from 95% to 40%
Example 2: User behavior
"Why aren't users completing onboarding?"
→ Observe: 60% drop-off at step 3
→ Hypothesize: H1: Form too long. H2: Unclear instructions. H3: Bug.
→ Test: Session recordings, form analytics, error logs
→ Result: H2 supported - users pause 30+ seconds at field X
Last Updated: 2026-02-02