MCP for developers · GA v1.0.0
Fractional cLevel’s executive knowledge, exposed as tools for your AI agent.
Self-service MCP server. Semantic search over the cLB methodology, SME diagnostics, product catalog. Bearer auth, OAuth 2.0 + PKCE, real rate limiting, granular scopes. Free tier available — register in 30 seconds.
What it is and what it’s for
Fractional cLevel is Yair Levy Wald’s external executive practice: a director covering 8 business dimensions (legal, infra, digital ecosystem, brand, team, AI governance, etc.) for SME owners.
Behind the practice is a Knowledge Engine: ~200 canonical documents (cLB Blueprints, cases, playbooks, decisions), vectorized, scope-gated. The MCP exposes that engine as tools your AI agents can call — Claude Desktop, Cursor, your own bot, whatever.
Real cases: Kontrol.ar uses it to enrich their internal bot with cLB methodology. External partners enrich SME-coaching chatbots with it.
Exposed tools
search_knowledgemcp.read.clbBúsqueda semántica sobre la metodología cLevel Blueprints (cLB). Recibe una pregunta, devuelve los docs más relevantes con BP, similarity y referencias.
query (string), bp_filter? (-1..5), max_results? (1..10)
analyze_clientmcp.read.clbRecibe contexto de un negocio (rubro, facturación, problema) y devuelve un diagnóstico estructurado: programa recomendado, BPs críticos, siguiente paso.
descripcion (string), rubro? (retail|servicios), facturacion_mensual? (string)
search_productsmcp.read.agenciaCatálogo de productos y programas de Fractional cLevel: LTM, ETM, LTE, ETE, Modelador, Bot de calificación, etc. Incluye precio y landing.
query? (string), line? (productos-ia|programas|servicios), pricing_model? (fixed|consultar), max_results? (1..50)
get_productmcp.read.agenciaFicha completa de un producto por SKU (descripción larga, todos los campos).
sku (string)
Tiers and rate limits
| Tier | Monthly | Per minute | Access | Notes |
|---|---|---|---|---|
| Free | 100 calls/mes | 10/min | Self-service | Para evaluación y prototipos. |
| User | 500 calls/mes | 30/min | Contactar | Uso personal de un dev. |
| Reseller | 10.000 calls/mes | 100/min | Contactar | Para partners que integran a sus clientes. |
| Enterprise | 50.000 calls/mes | 300/min | Contactar | Cuenta dedicada con SLA. |
Self-service grants mcp.read.clb scope (public methodology). For mcp.read.agencia or higher, contact yair@levywald.com.
How to connect (3 steps)
OAuth 2.0 with PKCE. Free tier, mcp.read.clb scope. The resulting access_token is an MCP API key — use it as Bearer against /mcp.
1) Register your app
curl -X POST https://www.fractionalclevel.ai/api/mcp/register \ -H "Content-Type: application/json" \ -d '{ "name": "Mi App", "email": "dev@miapp.com", "use_case": "Bot de coaching para PyMEs basado en cLB.", "redirect_uris": ["https://miapp.com/oauth/callback"] }' # Response: # { # "client_id": "mcpc_...", # "client_secret": "mcps_...", ← guardalo, no se vuelve a mostrar # "next_steps": { ... } # }2) OAuth authorize → code
GET https://www.fractionalclevel.ai/api/mcp/oauth/authorize? client_id=mcpc_...& redirect_uri=https://miapp.com/oauth/callback& response_type=code& scope=mcp.read.clb& code_challenge=<BASE64URL(SHA256(verifier))>& code_challenge_method=S256& state=<random> → 302 redirect a redirect_uri con ?code=<auth_code>&state=<...>3) Exchange code for access_token
curl -X POST https://www.fractionalclevel.ai/api/mcp/oauth/token \ -H "Content-Type: application/json" \ -d '{ "grant_type": "authorization_code", "code": "<auth_code>", "redirect_uri": "https://miapp.com/oauth/callback", "code_verifier": "<verifier>", "client_id": "mcpc_...", "client_secret": "mcps_..." }' # → { "access_token": "mk_...", "token_type": "Bearer", "scope": "mcp.read.clb", ... }
Examples
curl
curl https://www.fractionalclevel.ai/mcp \
-H "Authorization: Bearer mk_..." \
-H "Content-Type: application/json" \
-d '{
"jsonrpc": "2.0",
"method": "tools/call",
"params": {
"name": "search_knowledge",
"arguments": {
"query": "como armar el BP3 de marca y marketing",
"max_results": 3
}
},
"id": 1
}'JavaScript / TypeScript (MCP SDK)
import { Client } from '@modelcontextprotocol/sdk/client/index.js';
import { StreamableHTTPClientTransport } from '@modelcontextprotocol/sdk/client/streamableHttp.js';
const transport = new StreamableHTTPClientTransport(
new URL('https://www.fractionalclevel.ai/mcp'),
{ requestInit: { headers: { Authorization: `Bearer ${process.env.FC_MCP_KEY}` } } },
);
const client = new Client({ name: 'my-app', version: '1.0.0' });
await client.connect(transport);
const result = await client.callTool({
name: 'search_knowledge',
arguments: { query: 'BP3 marca y marketing', max_results: 3 },
});
console.log(result.content);Python (raw)
import os, requests
resp = requests.post(
"https://www.fractionalclevel.ai/mcp",
headers={
"Authorization": f"Bearer {os.environ['FC_MCP_KEY']}",
"Content-Type": "application/json",
},
json={
"jsonrpc": "2.0",
"method": "tools/call",
"params": {
"name": "search_knowledge",
"arguments": {"query": "BP3 marca y marketing", "max_results": 3},
},
"id": 1,
},
)
print(resp.json())Real use cases
Kontrol.ar
Enriches their internal coaching bot with cLB methodology. Scope mcp.read.kontrol — gets cLB + their own namespace.
Site bot
fractionalclevel.ai uses internal scope to power the lead-qualifier bot with the full base.
Doesn’t fit self-service?
For reseller / enterprise tier (incl. mcp.read.agencia with product catalog, mcp.read.all, etc.), reach out directly.
yair@levywald.comPublic manifest: /mcp · changelog: /api/mcp/changelog · register: POST /api/mcp/register