Job Posting Monitor MCP Server

MCP server for the Mamba Labs Job Posting Monitor actor on Apify.

Give it a list of role keywords and it returns the companies currently advertising those roles, one flat row per posting, each enriched with firmographics and the company LinkedIn URL.

Install

bash

npx -y @mambalabsdev/mcp-gtm-job-discovery

Claude Desktop

json

{
  "mcpServers": {
    "mamba-gtm-job-discovery": {
      "command": "npx",
      "args": ["-y", "@mambalabsdev/mcp-gtm-job-discovery"],
      "env": { "APIFY_TOKEN": "your-apify-token" }
    }
  }
}

Get an Apify token at console.apify.com/account/integrations.

Tool

`monitor_job_postings`

Give it a list of role keywords and it returns the companies currently advertising those roles, one flat row per posting, each enriched with firmographics and the company LinkedIn URL.

Input	Type	Required	Notes
`keywords`	array	yes	Role titles to search for, one per entry. Each keyword runs through the source chain separately.
`sources`	array	no	Job boards, run as an ordered chain. Tier 0, no key: `jobicy`, `remotive`, `arbeitnow`, `themuse`, `remoteok`, `weworkremotely`, `himalayas`. Tier 1: `adzuna` (needs `adzuna_app_id` and `adzuna_app_key`). Tier 2, only for keywords still short: `jobspipe`, `theirstack` (each needs its key) and `google_jobs` (needs `serpapi_key`). A posting on two boards is counted once. Default `["google_jobs"]`.
`country`	string	no	Geographic filter, for example United States, United Kingdom, Canada. Default `United States`.
`countries`	array	no	Loop these countries in one run. Overrides `country` when set.
`lookback_days`	integer	no	Only return postings newer than this many days. Default `30`.
`company_size_min`	integer	no	Drop companies with fewer employees than this (applied only when company size is known).
`company_size_max`	integer	no	Drop companies larger than this (applied only when company size is known).
`exclude_staffing_agencies`	boolean	no	Filter out staffing and recruitment agency postings. Default `true`.
`exclude_freelance_marketplaces`	boolean	no	Filter out postings where the company is a freelance marketplace (Upwork, Fiverr, and others). Default `true`.
`freelance_marketplaces`	array	no	Company names treated as freelance marketplaces. Default `["Upwork", "Fiverr", "Freelancer", "Toptal", "PeoplePerHour", "Guru", "99designs"]`.
`remote_only`	boolean	no	Only return remote postings. Default `false`.
`previous_run_date`	string	no	ISO date of your last run, for example 2026-06-01. Only newer postings are emitted, on top of the built in cross run delta cache.
`max_results`	integer	no	Upper bound on raw postings pulled before filtering. Higher values cost more. Default `100`.
`max_companies`	integer	no	Cap on unique companies enriched per run (firmographics and LinkedIn). Default `40`.
`serpapi_key`	string	no	Your SerpAPI key for the `google_jobs` source (free key at serpapi.com, 250 searches a month).
`adzuna_app_id`	string	no	Your Adzuna application id for the `adzuna` source.
`adzuna_app_key`	string	no	Your Adzuna application key for the `adzuna` source.
`jobspipe_key`	string	no	Your JobsPipe key for the `jobspipe` source. Billed by JobsPipe per job returned.
`theirstack_key`	string	no	Your TheirStack key for the `theirstack` source. Billed by TheirStack per job returned.
`metered_window_days`	integer	no	Ask the metered sources only for postings this many days old or newer, 1 to 7. Default `2`.
`enough_new_companies_per_keyword`	integer	no	Stop the chain for a keyword once this many new companies are found. Default `3`.
`metered_run_cap`	integer	no	Ceiling per metered source per country per run. Default `50`.

Keys passed as inputs travel in the tool call and stay in the chat transcript.

Billing

You are charged per result row returned, plus a small actor start fee. Filtered postings are free.

Pricing is on the actor's Apify page. Running this server consumes Apify credits.

What this server does and does not do

It is a thin client for the Apify actor. It passes your input through and returns the actor's output unchanged. Every behavior described above lives in the actor, not here.

Errors are surfaced, never swallowed. An invalid input, an invalid token, an exhausted balance, a timeout, or a run that returns anything other than a dataset all come back as an explicit tool error rather than as an empty result.