---
name: prototype
type: workflow
description: "Executes a rapid prototyping workflow that skips normal standards to quickly validate a product concept or core mechanic, producing throwaway code and a prototype report. Use when the user wants to quickly test an idea or mentions prototyping, proof of concept, or MVP spike."
argument-hint: "[concept-description]"
user-invocable: true
allowed-tools: Read, Glob, Grep, Write, Edit, Bash
effort: 3
when_to_use: "Use when the user wants to quickly validate a product concept or core mechanic before committing to full implementation, or mentions prototyping, proof of concept, or MVP spike."
When this skill is invoked:
Read the concept description from the argument. Identify the core
question this prototype must answer. If the concept is vague, state the
question explicitly before proceeding.
Read CLAUDE.md for project context and the current tech stack. Understand
what engine, language, and frameworks are in use so the prototype is built
with compatible tooling.
Create a prototype plan: Define in 3-5 bullet points what the minimum
viable prototype looks like. What is the core question? What is the absolute
minimum code needed to answer it? What can be skipped?
Create the prototype directory: prototypes/[concept-name]/ where
[concept-name] is a short, kebab-case identifier derived from the concept.
Implement the prototype in the isolated directory. Every file must begin
with:
// PROTOTYPE - NOT FOR PRODUCTION
// Question: [Core question being tested]
// Date: [Current date]
Standards are intentionally relaxed:
- Hardcode values freely
- Use placeholder assets
- Skip error handling
- Use the simplest approach that works
- Copy code rather than importing from production
Test the concept: Run the prototype. Observe behavior. Collect any
measurable data (frame times, interaction counts, feel assessments).
Generate the Prototype Report and save it to
prototypes/[concept-name]/REPORT.md:
## Prototype Report: [Concept Name]
### Hypothesis
[What we expected to be true -- the question we set out to answer]
### Approach
[What we built, how long it took, what shortcuts we took]
### Result
[What actually happened -- specific observations, not opinions]
### Metrics
[Any measurable data collected during testing]
- Frame time: [if relevant]
- Feel assessment: [subjective but specific -- "response felt sluggish at
200ms delay" not "felt bad"]
- User action counts: [if relevant]
- Iteration count: [how many attempts to get it working]
### Recommendation: [PROCEED / PIVOT / KILL]
[One paragraph explaining the recommendation with evidence]
### If Proceeding
[What needs to change for a production-quality implementation]
- Architecture requirements
- Performance targets
- Scope adjustments from the original design
- Estimated production effort
### If Pivoting
[What alternative direction the results suggest]
### If Killing
[Why this concept does not work and what we should do instead]
### Lessons Learned
[Discoveries that affect other systems or future work]
- Output a summary to the user with: the core question, the result, and
the recommendation. Link to the full report at
prototypes/[concept-name]/REPORT.md.
Important Constraints
- Prototype code must NEVER import from production source files
- Production code must NEVER import from prototype directories
- If the recommendation is PROCEED, the production implementation must be
written from scratch -- prototype code is not refactored into production
- Total prototype effort should be timeboxed to 1-3 days equivalent of work
- If the prototype scope starts growing, stop and reassess whether the
question can be simplified
Protocol
- Question: Reads concept from argument; if vague, states the core hypothesis explicitly before proceeding
- Options: Skip
- Decision: Prototype plan (3-5 bullets) shown and confirmed before implementation begins
- Draft: REPORT.md summary shown before saving
- Approval: "May I create
prototypes/[concept-name]/ and implement the prototype?"
Output
Deliver exactly:
- Prototype code in
prototypes/[concept-name]/ (throwaway — never imported by src/)
- Report saved to
prototypes/[concept-name]/REPORT.md
- Hypothesis verdict:
VALIDATED / INVALIDATED / INCONCLUSIVE
- Recommendation:
PROCEED (rewrite from scratch in src/) / PIVOT / ABANDON
1---2name: prototype3description: ---4---5---6name: prototype7type: workflow8description: "Executes a rapid prototyping workflow that skips normal standards to quickly validate a product concept or core mechanic, producing throwaway code and a prototype report. Use when the user wants to quickly test an idea or mentions prototyping, proof of concept, or MVP spike."9argument-hint: "[concept-description]"10user-invocable: true11allowed-tools: Read, Glob, Grep, Write, Edit, Bash12effort: 313when_to_use: "Use when the user wants to quickly validate a product concept or core mechanic before committing to full implementation, or mentions prototyping, proof of concept, or MVP spike."14---1516When this skill is invoked:17181. **Read the concept description** from the argument. Identify the core19 question this prototype must answer. If the concept is vague, state the20 question explicitly before proceeding.21222. **Read CLAUDE.md** for project context and the current tech stack. Understand23 what engine, language, and frameworks are in use so the prototype is built24 with compatible tooling.25263. **Create a prototype plan**: Define in 3-5 bullet points what the minimum27 viable prototype looks like. What is the core question? What is the absolute28 minimum code needed to answer it? What can be skipped?29304. **Create the prototype directory**: `prototypes/[concept-name]/` where31 `[concept-name]` is a short, kebab-case identifier derived from the concept.32335. **Implement the prototype** in the isolated directory. Every file must begin34 with:3536 ```js37 // PROTOTYPE - NOT FOR PRODUCTION38 // Question: [Core question being tested]39 // Date: [Current date]40 ```41 Standards are intentionally relaxed:42 - Hardcode values freely43 - Use placeholder assets44 - Skip error handling45 - Use the simplest approach that works46 - Copy code rather than importing from production47486. **Test the concept**: Run the prototype. Observe behavior. Collect any49 measurable data (frame times, interaction counts, feel assessments).50517. **Generate the Prototype Report** and save it to52 `prototypes/[concept-name]/REPORT.md`:5354```markdown55## Prototype Report: [Concept Name]5657### Hypothesis58[What we expected to be true -- the question we set out to answer]5960### Approach61[What we built, how long it took, what shortcuts we took]6263### Result64[What actually happened -- specific observations, not opinions]6566### Metrics67[Any measurable data collected during testing]68- Frame time: [if relevant]69- Feel assessment: [subjective but specific -- "response felt sluggish at70 200ms delay" not "felt bad"]71- User action counts: [if relevant]72- Iteration count: [how many attempts to get it working]7374### Recommendation: [PROCEED / PIVOT / KILL]7576[One paragraph explaining the recommendation with evidence]7778### If Proceeding79[What needs to change for a production-quality implementation]80- Architecture requirements81- Performance targets82- Scope adjustments from the original design83- Estimated production effort8485### If Pivoting86[What alternative direction the results suggest]8788### If Killing89[Why this concept does not work and what we should do instead]9091### Lessons Learned92[Discoveries that affect other systems or future work]93```94958. **Output a summary** to the user with: the core question, the result, and96 the recommendation. Link to the full report at97 `prototypes/[concept-name]/REPORT.md`.9899### Important Constraints100101- Prototype code must NEVER import from production source files102- Production code must NEVER import from prototype directories103- If the recommendation is PROCEED, the production implementation must be104 written from scratch -- prototype code is not refactored into production105- Total prototype effort should be timeboxed to 1-3 days equivalent of work106- If the prototype scope starts growing, stop and reassess whether the107 question can be simplified108109## Protocol110111- **Question**: Reads concept from argument; if vague, states the core hypothesis explicitly before proceeding112- **Options**: Skip113- **Decision**: Prototype plan (3-5 bullets) shown and confirmed before implementation begins114- **Draft**: REPORT.md summary shown before saving115- **Approval**: "May I create `prototypes/[concept-name]/` and implement the prototype?"116117## Output118119Deliver exactly:120121- **Prototype code** in `prototypes/[concept-name]/` (throwaway — never imported by `src/`)122- **Report** saved to `prototypes/[concept-name]/REPORT.md`123- **Hypothesis verdict**: `VALIDATED` / `INVALIDATED` / `INCONCLUSIVE`124- **Recommendation**: `PROCEED` (rewrite from scratch in `src/`) / `PIVOT` / `ABANDON`