Just Now MCP - Dynamic Creator Search Server

A Model Context Protocol (MCP) server built with Python and fastmcp supporting Streamable HTTP transport, STDIO transport, and Dynamic Influencer Attributes API Integration.

Features

🛠️ MCP Tools (1)

1. search_creator_facets (PostgreSQL / POSTGRES_DB_URL)
   • Formats and executes SQL queries against search_creator_facets(jsonb) functions.
   • Accepts dynamic attribute dictionaries (facets: dict[str, list[str]]).
   • Example query executed:

     SELECT *
     FROM search_creator_facets(
         '{
             "gender": ["Female"],
             "content_language": ["English"],
             "niche": ["Technology", "Gaming"],
             "country": ["United States"]
         }'::jsonb
     );
     

💡 MCP Skills / Prompts (1)

1. creator_attributes_skill
   • Live dynamic skill that fetches 37+ attribute schemas dynamically from ATTRIBUTES_API_URL (https://later-poc-api.aximsoft.com/api/v1/influencer_scout/attributes).
   • Automatically builds up-to-date Markdown schema tables detailing valid attribute codes (age_range, city, collaboration_type, content_language, country, niche, sub_niche, creator_tier, etc.) and their allowed option values for AI agents.

---

Environment Variables

Copy .env.example to .env:

# Dynamic Attributes API
ATTRIBUTES_API_URL=https://later-poc-api.aximsoft.com/api/v1/influencer_scout/attributes
VERIFY_SSL=False

# Database Connections
POSTGRES_DB_URL=postgresql://user:password@localhost:5432/creator_facets_db

# Server Configuration
MCP_SERVER_NAME="CreatorSearchMCP"
MCP_TRANSPORT="streamable-http"
MCP_HOST="0.0.0.0"
MCP_PORT=8000

---

Running the Server

Streamable HTTP Mode (Default)

uv run python main.py --transport streamable-http --port 8000

Docker Deployment

docker compose up --build -d