# AptiBuild AI — MCP Server

**Structured decision scaffolding for high-stakes career advice. For AI agents.**

Personality-conditioned idea generation, official BLS OEWS wage data, real-time FRED labor market signals, and pre-built harnesses tuned to specific life stages — without building and maintaining your own data pipelines. One endpoint, 12 tools, works with any MCP-compatible client.

[![Smithery Score](https://img.shields.io/badge/Smithery-98%2F100-violet)](https://smithery.ai/servers/parkertest196165/aptibuildai-mcp)

---

## Why this server, not public APIs

Raw BLS and FRED data are free to access — but packaging them for agent consumption isn't trivial. This server adds three things public APIs don't give you:

| What you get | Why it matters |
|---|---|
| **Personality-conditioned generation** | Most career APIs treat a job title as homogeneous. A risk-averse introvert nurse and a high-risk extrovert nurse get different outputs here. |
| **Pre-built harnesses** | Structured, multi-section intelligence documents built for agents to consume. One tool call gives your agent enough context to advise a human through a major career decision. Reduces hallucination on high-stakes advice. |
| **AptiBuild corpus** | 500+ app and business ideas sourced from real AptiBuild user profiles — not generic LLM speculation. |

Data layers: real-time FRED signals + BLS OEWS 2024 wages + AptiBuild proprietary corpus. US labor market data only.

---

## Connection

**MCP Server URL:** `https://mcp.aptibuildai.com/v1`

**Health check:** `https://mcp.aptibuildai.com/v1/health`

### Claude Desktop

Add to `claude_desktop_config.json`:

```json
{
  "mcpServers": {
    "aptibuild-ai": {
      "url": "https://mcp.aptibuildai.com/v1",
      "headers": {
        "X-API-Key": "YOUR_API_KEY_HERE"
      }
    }
  }
}
```

### Cursor

Add to `.cursor/mcp.json` in your project root:

```json
{
  "mcpServers": {
    "aptibuild-ai": {
      "url": "https://mcp.aptibuildai.com/v1",
      "headers": {
        "X-API-Key": "YOUR_API_KEY_HERE"
      }
    }
  }
}
```

### Replit Agent

In your Replit workspace → Agent panel → MCP settings → Add server:

```
Server URL: https://mcp.aptibuildai.com/v1
Header:     X-API-Key: YOUR_API_KEY_HERE
```

### Test without a subscription

```bash
npx @modelcontextprotocol/inspector https://mcp.aptibuildai.com/v1
```

Connect — `sample_career_pivot` is a free trial tool (10 calls, no key needed) so you can evaluate output quality before subscribing.

---

## Common Agent Workflows

These are the patterns agents use most:

**Career coach with unstructured input**
```
infer_aptitude_profile → generate_career_paths
```
User pastes a resume or LinkedIn bio. Agent infers the structured profile, then generates 4 personalized career paths.

**Side-hustle recommender**
```
infer_aptitude_profile → generate_app_ideas
```
Same two-step flow for business idea generation. Personality-conditioned — a risk-averse introvert and a high-risk extrovert in the same profession get different outputs.

**Quick profession lookup**
```
get_profession_profile
```
One call returns AptiBuild ideas + BLS wages for any profession. Best when you already know the job title.

**Harness + generation chain**
```
get_career_pivot_harness → generate_career_paths
```
Load structured context for the 38–55 cohort, then generate a personalized path on top of it. Reduces hallucination on high-stakes mid-career advice.

---

## Tools

### Free trial (no key required — 10 calls)

| Tool | What it does |
|---|---|
| `sample_career_pivot` | Returns the Transferable Skills section from the Career Pivot Harness for any profession. 1 of 5 sections — demonstrates data quality before subscribing. 10 free calls per IP, then a subscription key is required. |

### Paid (API key required — $39/month)

| Tool | What it does |
|---|---|
| `infer_aptitude_profile` ⭐ | **Start here for unstructured input.** Pass a bio, resume snippet, or LinkedIn summary — returns a structured aptitude profile ready to pipe into `generate_app_ideas` or `generate_career_paths`. Eliminates manual schema construction. |
| `get_profession_profile` | AptiBuild ideas + official BLS wages for a profession in one call. Best default for single-profession queries. |
| `get_career_pivot_harness` | 5-section intelligence for agents advising 38–55 year olds on career changes. Pivot success rates, transferable skill frameworks, high-fit industry pairs, live FRED signal pointers, 90-day action framework. |
| `get_second_act_harness` | 5-section encore career intelligence for 55+ professionals. Corrects BLS undercount of senior independent compensation. Covers fractional economy, FRED 55–64 and 65+ LFP signals. |
| `get_gig_economy_harness` | 5-section gig economy intelligence. The 1.5–2.5x earnings formula, growing vs. commoditized categories, platform landscape by category, FRED self-employment signals. |
| `generate_app_ideas` | Live AI — 4 custom business/app ideas conditioned on a full aptitude profile (skills, industry, 7 personality dimensions). ~5–15s. Harder to replicate with public data alone. |
| `generate_career_paths` | Live AI — 4 career path recommendations for the same profile. Returns transition paths, income potential, why each fits this specific person. ~5–15s. |
| `get_app_ideas_by_industry` | Up to 25 ideas for a given industry. Best for landscape scanning across a sector. |
| `get_app_ideas_by_profession` | Up to 25 ideas matched to a specific job title. |
| `lookup_occupation` | Resolves a job title to its BLS SOC occupation code. Use before `get_occupation_wages`. |
| `get_occupation_wages` | Official BLS OEWS 2024 wage data — median, percentile breakdown (10th/25th/75th/90th), total employment. Government-sourced, not estimated. |

---

## Quota & Rate Limits

| Tool group | Monthly quota | Typical latency |
|---|---|---|
| Lookup tools + `infer_aptitude_profile` | 500 / month | < 1s |
| Harness tools | 50 / month | < 1s |
| Live AI generation | 40 / month | 5–15s |

Hard cap: 100 calls / minute per key.

> **Note:** All labor market data (BLS OEWS, FRED) is US-based. The harnesses and generate tools draw from AptiBuild's proprietary corpus and work for any user regardless of location.

---

## Pricing

**Founding Integrator Rate: $39/month** — locks October 31, 2026

Price moves to $79/month on November 1, 2026. Subscribers who lock in before the deadline keep $39/month permanently.

Get your API key: [aptibuildai.com/developer](https://aptibuildai.com/developer)

---

## Enterprise

High-volume or custom quota needs? Email **aptibuildAI@proton.me** for enterprise terms.

---

## Data sources

- **FRED:** Federal Reserve Bank of St. Louis — [fred.stlouisfed.org](https://fred.stlouisfed.org)
- **BLS OEWS:** Bureau of Labor Statistics Occupational Employment and Wage Statistics, May 2024 release — [bls.gov/oes](https://www.bls.gov/oes/)
- **AptiBuild corpus:** Proprietary AI-generated dataset built and maintained by AptiBuild AI — sourced from real user skill profiles

---

## Part of the Parkland Consortium

AptiBuild AI is one of six independent builds in the [Parkland](https://parklandarchives.com) consortium — a family of products built with two means of entry: a front door for people, a back door for agents.

---

*Built on Replit. Made in Mississippi.*

---

**Last updated:** July 27, 2026
**MCP server status:** Live — `mcp.aptibuildai.com/v1`
**Pricing locked through:** October 31, 2026