Mentorly Is Now API-First (and Agent-Ready)

Everything you can do in Mentorly, you can now do through our API.

Today we are opening the same GraphQL API that the Mentorly application itself runs on. This is not a separate, limited "integration API" with a handful of endpoints. It is the real thing: every query and every mutation behind the product, from onboarding and matching to session booking, messaging, goals, reviews, and program analytics. If you can do it in the app, you (or software acting for you) can do it through the API.

The full reference lives at mentorly.com/developers. It covers authentication, rate limits, error handling, and every operation, and it is regenerated from our schema on every release, so it never drifts from reality.

Why this matters now: your AI agent can use Mentorly for you

The timing is not an accident. Teams everywhere are starting to delegate routine work to AI agents, and we built this API so those agents can work inside your mentoring program the same way you do.

An API key in Mentorly is bound to a person and acts as that person. That one design decision does a lot of work:

• A program manager's agent can pull the live roster, check which mentees still have no match, read program analytics, and prepare the monthly report from real session data.
• A mentor's agent can publish availability, accept incoming session requests, and reply to messages.
• A mentee's agent can book a session inside a mentor's real availability, send the prep message, and track goals through to completion.

Because the key acts as the user, the agent inherits exactly that user's permissions and nothing more. There is no separate permission model to configure and no way for an agent to see data its owner could not see in the app.

Built for trust, not just for demos

Opening an API is easy. Opening it responsibly is the actual work, and most of what we shipped is governance:

• Scopes are a ceiling. Every key carries scopes such as read:bookings or write:bookings. A read-only key cannot change anything, even though its owner could in the app. When an agent oversteps, the API answers with a clear, machine-readable refusal.
• Agents are visible in the audit trail. Every change records whether the person did it themselves or their agent did it through a key, and which key. "Who did this?" always has a real answer.
• Every key has its own rate limit. Your integration gets its own quota with standard rate-limit headers, so one runaway script never competes with the people using the app.
• Keys are revocable in one click, shown once at creation, and stored only as a hash on our side.

For the humans: integrations and answers

Agents are the headline, but the same API serves plainer needs. Pull your program's data into your own BI tools. Sync members from your HR systems. Build the internal dashboard your leadership team keeps asking for. Anything the Mentorly dashboard can show you, the API can feed into the tools you already use.

And for developers specifically: the API is fully documented at mentorly.com/developers, with authentication, pagination, error codes, and example patterns, written to be equally readable by a developer and by an AI agent pointed at it.

How to get access

API keys are issued by our team and bound to a user in your program. Talk to your Mentorly account manager or write to support@mentorly.co, tell us what you want to build, and we will set you up with a key scoped to exactly that.

We think mentorship software should work the way the rest of your stack works: open, documented, and ready for whatever is doing the work on your side of the API, human or agent. As of today, Mentorly does.