Workflows

Talent scout

Run a local scouting workflow that searches Markidy profiles and evaluates each person against your scouting goal using their public profile introduction, career history, recruiting preferences, social links, and trust links.

Watch demo

See the full flow from setup to saved scouting report before installing the workflow.

Open on YouTube

Overview

Talent scout is an open-source local app for recruiters, founders, and hiring managers. It uses the Markidy API to find public profiles, then reviews each candidate against your scouting goal using structured profile signals, including bio, career history, recruiting preferences, social links, and trust links.

Talent scout does not just match keywords. It evaluates public profile signals such as a person's introduction, career history, recruiting preferences, social links, and trust links against your scouting goal.

The workflow is designed for sourcing assistance. It does not send messages, run chats, or make automated employment decisions.

Find profile matches Searches Markidy profiles, then reviews richer public profile signals beyond simple keywords.
Evaluate fit Uses an AI provider to compare your requirements with each profile's bio, career, conditions, social links, and trust links.
Save reports Stores completed reports locally and supports JSON or CSV export.

Requirements

ItemRequiredNotes
Node.jsYesUse Node.js 20.19+ or 22.12+.
Markidy API keyYesCreate one from API Keys.
AI provider keyYesOpenAI, Anthropic Claude, Google Gemini, xAI Grok, or Mistral AI.

Quick start

Clone the repository, then use the launcher for your OS or run the terminal commands manually.

Clone
git clone https://github.com/Team-Daco-V3/talent-scout.git
cd talent-scout

Option 1: launcher

Windows
start-talent-scout-windows.cmd
macOS or Linux
chmod +x start-talent-scout-macos-linux.sh
./start-talent-scout-macos-linux.sh

Option 2: terminal

npm
npm install
npm run start

On Windows PowerShell, use npm.cmd if script execution blocks the regular npm command.

Windows PowerShell
npm.cmd install
npm.cmd run start

The local app starts at http://localhost:17300. Stop the server with Ctrl+C in the terminal.

First run

  1. Open http://localhost:17300.
  2. Keep the default Markidy API URL: https://api.markidy.com.
  3. Paste your Markidy API key.
  4. Choose an AI provider and paste that provider's API key.
  5. Create or edit a scout with the goal, target role, required skills, location or work conditions, and candidate count.
  6. Save the scout, then run it.
  7. Open Reports to review saved and in-progress results.

Reports and security

Completed runs are saved as JSON files in the local project reports folder. Existing reports are not overwritten, and API keys are not written into report files.

Hosted deployment warning Treat hosted versions differently from local use. Add a backend trust model, rate limits, request logging policy, and proper secret handling before accepting user-provided API keys.