Connect Qalentis to your AI assistant
Assistants such as Claude, ChatGPT and Cursor connect to Qalentis through the Model Context Protocol (MCP). Ask "summarise the applicants for this job" or "show my team's OKR status" and the assistant answers from your Qalentis data, within the permissions you grant.
Connection URL
Use this URL when you add a new connector in your assistant.
https://mcp.qalentis.com/mcp
How to connect
First create an API key in Qalentis: Settings → API keys. The default is "read-only"; choose "full access" if the assistant should create records. API keys page →
Claude
- Settings → Connectors → "Add custom connector".
- Name it "Qalentis" and paste the connection URL.
- Click "Connect" and paste your API key into the Qalentis page that opens.
ChatGPT
- Settings → Connectors → create a new connector (developer mode must be on).
- Enter the connection URL and choose OAuth for authentication.
- Paste your API key into the Qalentis page that opens.
Cursor and others
- Add a new "Streamable HTTP" server in the MCP settings.
- URL: the connection URL; header: Authorization: Bearer qali_live_…
- To load only the groups you need, add ?scope=hiring,ats to the URL.
Safeguards
- Your key is never shown to the assistant itself; with OAuth the assistant only receives a short-lived token.
- New keys are read-only by default. Even with a full-access key, decisions and deletions are not applied until an authorised person at your company approves them in Qalentis.
- Each key only sees its own company's data; companies are also separated at database level.
- Every call made through the assistant is logged, and you can revoke the key in Qalentis at any time.
- Data the assistant reads goes to the AI provider you chose (e.g. Anthropic, OpenAI). You decide which assistant and which tool groups to enable.
Tools
111 tools in 5 groups. The assistant picks the right tool for your question.
Core: profile, company summary, notifications — always on13
Assessment: tests, interviews, invitation links, candidate results42
ATS: job postings, applications, candidate pool, CV analysis18
Performance: competencies, reviews, 9-Box, OKR, 360°, potential32
Talent: skills profile, role readiness, onboarding6
