Mulyra
Real-time salary data inside your AI editor.
Connect Claude Desktop, Cursor, or other agents directly to Mulyra's database of 2M+ job postings using the Model Context Protocol (MCP).
Quick Configuration
Mulyra hosts the server, so there is nothing to install and no API key needed. Point your assistant at the URL below.
Claude (desktop or web)
Settings > Connectors > Add custom connector, then paste this URL
https://api.mulyra.com/mcp
Claude Code
Run in your terminal
claude mcp add --transport http mulyra https://api.mulyra.com/mcp
Cursor, Windsurf, VS Code
Add to the mcpServers section of your MCP config (e.g. ~/.cursor/mcp.json)
{ "mcpServers": { "mulyra": { "url": "https://api.mulyra.com/mcp" } } }
Supported Tools
Once connected, your AI assistant can invoke any of the following capabilities autonomously during conversations.
search_jobs
Search job listings from Mulyra real-time database. Returns up to 20 matching jobs with salary and location data.
Arguments:
query (optional), location (optional), min_salary (optional), max_salary (optional)
Returns:
Count and list of matching job objects with title, company, salary ranges, and URLs.
get_salary_benchmark
Get salary statistics (p10, median, p90) for a role from active job postings.
Arguments:
role (required), location (optional city), country (optional, default US)
Returns:
p10, median, and p90 annual base pay in local currency, based on active postings with disclosed pay ranges.
compare_offer
Compare two job offers side-by-side. Computes estimated annual total compensation and returns a structured recommendation.
Arguments:
offer1 (required), offer2 (required)
Returns:
Estimated annual total compensation for each, differences, and a structured recommendation.
get_company_intel
Get company info: active job count, salary range, and top open roles.
Arguments:
company (required)
Returns:
Company active roles count, salary range (min, average, max), and top open positions.
add_to_tracker
Save a job to the authenticated user's tracker. Requires a valid Mulyra Bearer token in the Authorization header.
Arguments:
job_id (required)
Returns:
Status confirmation string indicating success or if the job is already tracked.
get_top_skills_for_role
What skills are employers requiring most for this role right now?
Arguments:
title (required), country (optional)
Returns:
Top 10 skills required by employers for this role, with frequency counts and sample size.
compare_salary_relocation
Side-by-side salary comparison across two countries (real market data).
Arguments:
title (required), from_country (required), to_country (required)
Returns:
Median salary, local currency, USD equivalent, and cost-of-living index for origin and destination.
get_hiring_velocity
Is demand for a role up or down in a market? Returns whether demand is rising or falling.
Arguments:
title (required), country (optional)
Returns:
Current and prior 30-day posting counts, percentage change, and up/down/stable trend.
get_days_to_fill
How long do roles like this typically stay open?
Arguments:
title (required), country (optional)
Returns:
Typical days a role stays open, with sample size.
get_salary_at_close
What did companies actually pay for this role? Returns a pay percentile distribution for recently filled roles.
Arguments:
title (required), country (optional), period (optional, days, max 180)
Returns:
Salary percentile distribution (p10-p90) from recently-filled postings.
get_fill_rate
What percentage of posted roles actually get filled?
Arguments:
country (optional)
Returns:
Fill rate for a country.
get_salary_trend
Are salaries for this role going up or down?
Arguments:
title (required), country (optional)
Returns:
Recent monthly salary figures and the trend direction.
Get your Developer Token
Authenticate your AI assistant to track jobs, monitor application status, and unlock high-limit benchmarks.
Frequently Asked Questions
What is the Model Context Protocol (MCP)?
MCP is an open standard designed by Anthropic that allows AI applications like Claude Desktop, Cursor, or IDE extensions to securely connect to external APIs and databases. Rather than writing custom plugins for each LLM, MCP provides a standard way to expose tools and resources to any assistant.
How do I authenticate tools that require a Mulyra token?
Public tools like search_jobs, get_salary_benchmark, compare_offer, get_company_intel, get_top_skills_for_role, compare_salary_relocation, get_hiring_velocity, get_days_to_fill, get_salary_at_close, get_fill_rate, and get_salary_trend require no authentication. The add_to_tracker tool saves jobs to your Mulyra account, so it needs a signed-in Mulyra session token sent as a Bearer token; developer API keys are not accepted for it.
Are there rate limits on the MCP server?
Yes. The MCP server allows 30 requests per minute per IP address, and answers to identical questions are cached for up to an hour. Interactive use stays well below the limit; automated bulk requests may be throttled.

MCP Server for AI Assistants

Connect your AI assistants (Claude Desktop, Cursor, Copilot) directly to Mulyra's live salary database and job search using the Model Context Protocol.

Salary data by role ยท Open jobs