StaffPick MCP¶

A job post is a guess about the market. The number that decides whether the hire works is what the same role actually pays somewhere else.

Server type: Remote (Streamable HTTP), two endpoints
Search:      https://staffpick.ai/api/mcp (no sign-in, 5 tools)
Full:        https://staffpick.ai/mcp (OAuth 2.1 with PKCE)
Server:      staffpick 2.2.0
Tools:       5 at tools/list on the search endpoint
Coverage:    Remote workers across support, bookkeeping, development and more

Why This Matters for Operators¶

Hiring a remote assistant or bookkeeper is the first operational hire most small businesses make and the one with the least reliable market information. Rates quoted by an agency, a rate seen in a forum and a rate a competitor pays are three different numbers, and none of them is attached to a person who was tested.

This server turns three real questions into tools: who fits a role and what did they actually score, what does this role pay at these hours, and what should the brief say. salary_benchmark returns the 25th to 75th percentile of monthly pay for a role at the given weekly hours, computed from the rates of approved profiles on the platform rather than from a survey.

It is a different slice from what the catalogue already holds. Ruzora places LATAM engineers with compliance built in, Ashton & Forge lists vetted AI implementation agencies and the lemon-io entry hires senior developers. None of them scores candidates on a graded test and prices the role at the same time.

What It Returns, Verified¶

A bare initialize on https://staffpick.ai/api/mcp returned HTTP 200 with no credential as staffpick 2.2.0, declaring the MCP Apps UI extension, and tools/list returned five tools:

search_talent searches tested remote workers across virtual assistants, support, bookkeeping, developers and more, and every result carries the test scores taken on StaffPick rather than self-reported skills.
get_talent returns one anonymised profile: summary, skills, tools, languages, availability, the business hours covered, rates, and test results with dates, including a CEFR band for English.
salary_benchmark returns typical monthly pay in USD for a role at given weekly hours, as the 25th to 75th percentile from real approved profiles, and says not enough data rather than inventing a figure.
create_role_brief takes the hiring request in the operator's own words, for example part-time bookkeeper who knows Xero, mornings in Sydney, under $1,000 a month, and turns it into a structured brief.
request_intro returns a StaffPick link the human opens to request an introduction. It deliberately never returns contact details, so a conversation with a candidate cannot start without the operator acting.

Limits Worth Recording¶

The search endpoint is read-only and anonymous, and it is the surface this entry verified. The signed-in endpoint at https://staffpick.ai/mcp requires OAuth 2.1 with PKCE, answered the probe with HTTP 401 and a resource_metadata pointer, and is where the platform's own actions live. Candidacy rates, availability and scores describe the platform's own pool, so the benchmark is a real market signal and not the whole market. request_intro means the platform owns the introduction, which is the opposite of a lead list and should be read that way before someone expects contact details.

FAQ¶

Do I need an account to try it?¶

No. The five search and benchmark tools answer without a credential, which is how this entry was verified.

Where do the salary numbers come from?¶

The rates of real approved StaffPick profiles, returned as a 25th to 75th percentile for a role at a given number of weekly hours in USD, with an explicit not-enough-data answer when the sample is thin.

Does it hand over candidate contact details?¶

No. request_intro returns a link the operator opens to request an introduction, and the tool description states it never returns contact details.

Is this a recruitment agency?¶

It is a marketplace of tested remote workers with a search and intro surface exposed to agents. The testing and the intro are the platform's, and the hiring decision stays with the operator.

How is this different from the other hiring entries here?¶

Ruzora is LATAM engineering placement with compliance, Ashton & Forge is a directory of AI implementation agencies and lemon-io targets senior developers. This one carries graded test scores and a salary benchmark for general remote roles.

Integration with CorpusIQ¶

CorpusIQ is the connector layer that lets AI assistants read the systems a business already runs on - accounting, payments, CRM, email, files and analytics - and answer with the source named. It is read-only by design: it retrieves data and never writes to, or acts in, a connected system, which is what makes it safe to point at live accounts.

This page is part of the CorpusIQ directory, the reference we maintain for the MCP server and agent-skill ecosystem. Run alongside CorpusIQ, a third-party server like this one can be combined with a company's own business data in a single answer.

See what CorpusIQ connects