---

name: show-assessment description: Show a real r2r assessment example. Displays what an actual assessment output looks like from the Research-to-Roadmap plugin.

Show Assessment Example

When triggered, display this real assessment output:

---

Assessment: Caching Layer POC

Generated by r2r:assess

Executive Summary

What Was Proven

✅ Validated Claims:

1. Redis caching reduces DB load by 60%

   • Evidence: benchmark-results.md shows 847 → 339 queries/sec
   • Conditions: Tested with 10K concurrent users
   • Confidence: High

2. Cache invalidation strategy works

   • Evidence: No stale data in 48-hour soak test
   • Conditions: Write-through with TTL backup
   • Confidence: High

3. P95 latency improved from 230ms → 45ms

   • Evidence: Load test results
   • Conditions: Warm cache state
   • Confidence: High

What Was Assumed (Not Proven)

⚠️ Unvalidated Assumptions:

1. Multi-region replication will work

   • Risk: High - not tested
   • Recommendation: Spike needed before H2

2. Cost projection at 10x scale

   • Risk: Medium - extrapolated only
   • Recommendation: Build cost model

3. Cache warming strategy for cold starts

   • Risk: Medium - not addressed
   • Recommendation: Add to M1 requirements

Gap Analysis

Technical Debt Identified

CopyDEBT INVENTORY:

🔴 Critical:
   - Hardcoded Redis connection strings
   - No circuit breaker for Redis failures

🟡 Important:
   - Cache key generation not standardized
   - No cache size limits configured

🟢 Minor:
   - Test coverage at 67% (target: 80%)
   - Some TODO comments remain

Risks & Blockers

Recommendation

PROCEED WITH CAUTION

The POC demonstrates clear value (60% DB load reduction, 5x latency improvement). However, production deployment requires:

1. Before H1: Circuit breaker, connection string externalization
2. Before H2: Multi-region testing, cost modeling
3. Before GA: Full monitoring suite, runbook documentation

---

Run /r2r:decompose to break this into shippable components.

---

This is a representative example. Actual assessments are generated from your research artifacts.