Replicability Scorer: Tech Stack Fit Analysis
Replicability Scorer: Tech Stack Fit Analysis
The Replicability Scorer is a scoring sub-module that assesses how easily any discovered product could be rebuilt using the AgentOS stack: Next.js, Neon, Inngest, tRPC, Drizzle, and Vercel.
This score is one of the four pillars that make up a product's composite opportunity score on the dashboard.
How It Works
When a product is deep-analysed, the system passes the product's extracted feature list to the LLM with a structured prompt. The LLM evaluates the product across four dimensions:
1. Data Model Complexity
How intricate are the underlying data structures likely to be? A simple CRUD SaaS with a handful of entities scores high. A product with complex relational schemas, multi-tenancy requirements, or heavily nested hierarchical data scores lower.
2. Real-Time Requirements
Does the product depend on live data feeds, websocket connections, or sub-second latency? Products that are primarily request/response in nature are a better fit for the stack and score higher on this dimension.
3. Third-Party Integration Count
How many external services or APIs does the product rely on? High integration counts increase build complexity and ongoing maintenance burden. Products with few or zero critical external dependencies score higher.
4. AI Component Needs
Is AI/ML central to the product's core value proposition? Products that use AI as a thin layer over structured data are easier to replicate than those requiring custom model training, specialised embeddings, or complex inference pipelines.
Output
The scorer outputs two things:
- A replicability score (1–10) — a single numeric signal used in composite ranking
- A written rationale — a short LLM-generated explanation of the score, visible inside the product dossier
Score Reference
| Score | Signal |
|---|---|
| 8–10 | Strong stack fit — could be built with standard AgentOS patterns with minimal friction |
| 5–7 | Moderate fit — some complexity areas exist but nothing that blocks delivery |
| 1–4 | Weak fit — significant real-time demands, integration sprawl, or specialised AI needs make this a heavy lift |
Where to Find It
- Dashboard — The replicability score column is included in the sortable product rankings table. Click the column header to sort by replicability.
- Product Dossier — Open any product's full dossier to read the LLM-written rationale beneath the score, alongside the breakdown by dimension.
Relationship to Composite Score
The replicability score is one component of the overall opportunity score used to rank products. A product with high market demand but a replicability score of 2 will rank lower overall than a product with slightly lower demand but a score of 9 — because build feasibility directly affects real opportunity value.
See the Opportunity Scoring page for the full composite scoring formula.
Notes
- Replicability scoring runs automatically during deep analysis. No additional configuration is required.
- Scores are generated fresh each time a product is re-analysed; they are not cached across crawl sessions.
- The score reflects the AgentOS stack specifically — it is not a general "build difficulty" metric.