Model Context Protocol

Connect AI assistants to SkillCertify

SkillCertify runs a Model Context Protocol (MCP) server so AI assistants like Claude and ChatGPT can work with your hiring data on your organization's behalf — search verified talent, run skills assessments, manage job postings and read results. Every sensitive or public action is confirmed by a person on SkillCertify.

MCP endpoint
https://skillcertify.io/v1/mcp
Transport: JSON-RPC over HTTP · Auth: OAuth 2.1 with PKCE · Scopes: read, assess, jobs, community
Create an organization

What an AI assistant can do

Read (any team member)
scope: read
  • Search skills tests and hands-on labs
  • Search verified talent by skill and score
  • Read candidate progress, results and workflow analytics
  • List job postings and their applicants
  • View your team, credit balance and action status
  • Read community threads and replies
Assess (owner / admin)
scope: assess
  • Create and edit assessment workflows
  • Invite candidates (confirmed by a person)
  • Add candidates to your Talents Pipeline
  • Cancel an assessment
Jobs (owner / admin)
scope: jobs
  • Create job postings as drafts
  • Publish or edit a job (publishing is confirmed by a person)
Community (owner / admin)
scope: community
  • Post threads and replies under your organization's name (confirmed by a person)

How to connect

  1. Create or sign in to a SkillCertify organization account.
  2. In your MCP-compatible AI client, add the endpoint https://skillcertify.io/v1/mcp.
  3. Sign in and authorize the scopes you want. Your role decides what's available (members get read-only).
  4. The assistant can now work with your hiring data. Inviting candidates, publishing jobs and posting to the community always wait for your confirmation on SkillCertify.

Frequently asked questions

Does SkillCertify have an MCP server?
Yes. SkillCertify runs a Model Context Protocol (MCP) server for organizations at https://skillcertify.io/v1/mcp. AI assistants that support MCP can connect to it with your permission to work with your hiring data.
Which AI assistants can connect?
Any MCP-compatible client — for example Claude (Claude Desktop and Claude.ai), and other agents or IDEs that support the Model Context Protocol. You add the endpoint https://skillcertify.io/v1/mcp and sign in with your SkillCertify organization account.
What can an AI assistant do through the MCP server?
Read-only: search skills tests and verified talent, read candidate progress, results and analytics, job postings and applicants, your team and credit balance. With permission it can also create and manage assessment workflows, invite candidates, create and publish job postings, and post in the community forum under your organization's name.
Is it secure?
Yes. Access uses OAuth 2.1 with PKCE and scoped permissions (read, assess, jobs, community) tied to the user's role. Every sensitive or public action — inviting candidates, publishing a job, posting to the community — must be confirmed by a person on SkillCertify before it happens. Candidate-authored text is treated as data, never as instructions.
How do I connect?
Create or sign in to a SkillCertify organization account, add the MCP endpoint https://skillcertify.io/v1/mcp in your AI client, authorize the scopes you want, and the assistant can start working with your hiring data.