validate-hypothesis¶
Tests design assumptions through code experiments, codebase analysis, and external research.
Synopsis¶
Description¶
The validate-hypothesis command tests assumptions made in your technical design before implementation begins. It runs experiments, analyzes the codebase, and researches external sources to validate (or invalidate) design decisions.
This is an optional step in the feature workflow, recommended when:
- The design includes unfamiliar technologies
- Performance assumptions need validation
- Integration approaches are uncertain
- Risk mitigation is important
Parameters¶
| Parameter | Position | Required | Default | Description |
|---|---|---|---|---|
FEATURE_ID |
$1 |
Yes | - | Feature identifier |
Prerequisites¶
design.mdmust exist in.rp1/work/features/<feature-id>/
Output¶
Location: .rp1/work/features/<feature-id>/hypothesis-results.md
Contents:
- Hypotheses extracted from design
- Validation approach for each
- Experiment results
- Codebase analysis findings
- External research summary
- Recommendation (proceed/modify/reconsider)
Examples¶
Validate Design Assumptions¶
Example output:
✅ Hypothesis Validation Complete
Hypotheses Tested: 3
1. "Redis can handle 10k requests/second"
Status: VALIDATED
Evidence: Benchmark showed 15k req/s
2. "Existing middleware supports caching headers"
Status: VALIDATED
Evidence: Found in src/middleware/cache.ts
3. "Cache invalidation can use pub/sub"
Status: NEEDS_MODIFICATION
Evidence: Current setup uses polling
Recommendation: Update design to use polling pattern
Overall: PROCEED with modifications
Validation Methods¶
| Method | Use Case |
|---|---|
| Code experiments | Performance, API behavior |
| Codebase analysis | Pattern compatibility, dependencies |
| External research | Library capabilities, best practices |
Related Commands¶
feature-design- Creates the design to validatefeature-tasks- Next step after validation