Mentoring programs from your AI assistant: find mentors, book meetings, keep notes and next steps.
Scored across 22 tools
Most tools target a distinct resource+action pair, and the detailed descriptions help separate close pairs like get_mentoring_tool (one assigned worksheet) vs list_mentoring_tools (assigned worksheets) vs list_tool_catalog (available worksheets). The 'what needs doing' cluster (get_my_todo, list_tasks, list_next_steps) overlaps conceptually, though the descriptions draw the lines reasonably well.
Every tool follows a clean verb_noun snake_case pattern (find_, get_, list_, read_, search_), with a consistent 'my' / 'mentoring' qualifier scheme for personal vs catalogue resources. No convention mixing.
22 tools is on the heavy side but the domain is genuinely broad (programmes, meetings, slots, profiles, matching, worksheets, forms, posts, tasks, conversations). Each tool maps to a real sub-resource; a couple of the listing tools could be consolidated but none look gratuitous.
Read coverage is excellent (programmes, meetings, profiles, matches, forms, conversations, tasks), but the surface is essentially read-only: described actions like booking a slot via `start`, ticking off task items, submitting form answers, or sending a worksheet have no corresponding write tool. Agents will hit dead ends when the user asks to actually do something rather than look at it.
Free times in a mentor's published hours that the person (a mentee) can book, after the mentor's and the person's own calendars, blocked times and the programme's minimum notice. Each free time has the start that books it. As mentor, with mentor_id empty: the person's own free published times, also free in the calendar of the mentee (or group) in with_member_id.
Inputs: program (required), days, from, mentor_id, +1 more. Returns: from, note, slots, until, +2 more
The questions of a programme form or survey the person was asked to fill, with each question's id and the form's link in the web app.
Inputs: form_id (required), program (required).
One meeting: time, place or video link, who, its goal in a structured programme, whether it can still be cancelled, the meeting notes the person may read (their own and the shared ones), and the before and after meeting forms of a structured programme: the person's answers and, where the programme shares them, the other side's.
Inputs: meeting_id (required).
The profile of a mentor, a mentee or a mentee group (team, startup, venture) in a programme, as the person may see it in the web app: headline, about, and the answers to the programme's profile questions; for a group also its members, website and LinkedIn, which are what to look the venture up by on the web.
Inputs: program (required), member_id (required), role.
One assigned mentoring tool: what each section asks, the answer format each section takes, and the answers so far. Some visual sections can only be filled in the web app.
Inputs: program (required), assignment_id (required).
The person's profile in a programme: headline and bio, their answers to the programme's profile questions (with question ids and which ones they may change), where mentees are teams or startups, their team's profile, and which calendar and video accounts (Google, Microsoft, Zoom) they connected and whether each still works.
Inputs: program (required).
Everything waiting for the person across all programmes: upcoming meetings, meeting forms to fill, match requests to answer or waiting, programme to-do lists, surveys, mentoring tools assigned, and programme events. Start here when the person asks what to do next.
Inputs: no named parameters. Returns: not_shown, todo_lists, forms_to_fill, match_requests, +5 more
Explains how one programme works for the person: their role, how matching works (admin-matched, mentee selects, open pool, network), structured or freestyle, who can book meetings and how, limits (meetings per week, cancellation notice), meeting goals of a structured programme, the programme's own words for mentor/mentee/group, and its time zone.
Inputs: program (required).
The person's message threads, newest first, with the last message and whether it is unread. Optionally only the threads with one member (with_member_id) or in one programme.
Inputs: program, unread_only, with_member_id.
The person's meetings: upcoming (soonest first) or past (newest first) in one programme, or, with from, every meeting on the given days ("who am I meeting today?") in one programme or, without program, in all of them. Each meeting lists the people or group it is with, by member id and role, for reading their profiles.
Inputs: days, from, when, program, +1 more. Returns: from, note, count, until, +2 more
Mentoring tools (worksheets) assigned in a programme: to the person as mentee, or by the person as mentor, with status, so a mentor sees who handed in what.
Inputs: program (required), tool, status.
The hours the person published for meetings (as mentor, or as mentee where the programme lets mentees publish hours): each one with its repeat, length of a meeting, place and the periods they closed, plus the times still free in the asked days. Across all programmes unless program is given.
Inputs: days, from, program.
The person's mentor/mentee pairs in a programme: active matches and requests sent or received, with the match id other tools take.
Inputs: program (required). Returns: matches, program, statuses
Lists the mentoring programmes the person belongs to as mentor and/or mentee, with the id/domain other tools take as program. Programmes whose admins turned off AI assistants are listed with ai_access=false and cannot be used.
Inputs: no named parameters. Returns: count, programs
The next steps of the person's matches in a programme: what is to be done, who does it, by when, and whether it is done. Open ones first, overdue at the top. mine keeps only the ones the person does.
Inputs: program (required), mine, match_id.
Announcements the programme posted for the person's role, newest first, with comment counts. Comments are read with include_comments_of.
Inputs: program (required), role, include_comments_of.
The programme's to-do lists for the person, item by item, with what is done. Items can be ticked off one by one.
Inputs: program (required).
The mentoring tools (worksheets) this programme offers, with what each is for and how long it takes. A mentor can send any of them to matched mentees.
Inputs: program (required).
The latest messages of one message thread, oldest first. Reading here does not mark them read.
Inputs: conversation_id (required), limit.
Searches the mentees (or, where mentees are groups, the groups) the person may see in a programme. Returns each one's member id.
Inputs: program (required), page, query.
Searches the mentors the person may see in a programme (in a programme where mentees choose, only those they may request), for example by a skill ("sales"); with free_on, only those with bookable free times then. Returns each mentor's member id.
Inputs: program (required), days, page, query, +1 more.
The ventures (startups, teams: the programme's groups) on the programme's ventures page, as the person may see them there; refused where the programme does not show that page to the person. Narrows by words in the name, description and profile answers (query), and by one profile question and its answer (question + answer, for example question "sector" and answer "edtech"). Each venture comes with its venture_id, description, website, its answers to the programme's profile questions and, for a mentor, how many meetings the person has held with it and when the last one was; only_met lists just the ventures the person has met. The first page also lists the profile questions that can narrow the list, with their options. Members, LinkedIn and the full answers are in the venture's profile (role group).
Inputs: program (required), page, sort, query, +3 more.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Employee mentoring software. Goals, meetings, messages and surveys from your AI assistant.
Prep 1:1s, draft feedback and work through hard conversations at work, in your AI.
AI mentoring fit and book: check the fit with Marian Kamenistak and book a session in 10 minutes.
AI executive assistant: email triage, inbox, calendar, files, memory, phone and connected apps.
Add one secure layer between your agents and this server.
