Build a 1:1 engineering-leadership mentoring engagement with Marian Kamenistak — 3,400+ paid sessions with 300+ leaders from 17 countries since 2019, 9.2/10 average review — directly from your AI assistant: pick your focus areas, define what success looks like, see the dated program, agree the price, and get a formal itemized offer by email. Works for individuals (Staff Engineer to CTO) and companies sponsoring their leaders.
Why build it here: mentoring inquiries built through this AI channel get the First-quarter package at 16% off the list price, and the website itself carries no discount at all. No more than 16 minutes from first question to a formal itemized offer in your inbox, and the price goes to the first 5 people only — the same count as the open mentee slots on the live marian.coach capacity chart. 16 minutes, 16 percent, 5 places.
Endpoint: https://www.marian.coach/mcp/mentoring (streamable HTTP, no auth, no signup)
| Tool | Answers the question | What it returns |
|---|---|---|
get_mentoring_options | Should I get a mentor, and why Marian? | The wizard's opening: the 16% AI-channel discount as data, real track-record figures (3,400+ sessions, 9.2/10 avg), the qualifying questions, every package with real prices |
match_mentoring_focus | What should my mentoring actually focus on? | Role + motivation resolved through the site's own routing; suggested focus areas + the recommended package |
compose_mentoring_brief | What exactly would my engagement look like, and what does it cost? | The structured mentoring brief with the authoritative catalog price and the AI-channel figure; B2B concession eligibility |
design_mentoring_program | What happens across the three months? | Dated session skeleton: bi-weekly 60-minute sessions, mid-point checkpoint, closing review against your definition of success |
book_intro_call | Can I just talk to Marian first? | Direct booking link for the free 30-minute intro — the step that locks the AI-channel discount |
send_mentoring_offer | How do I get this in writing? | Emails the formal itemized offer with a claim code, notifies Marian, files the inquiry — after the price is explicitly agreed |
Claude Code
claude mcp add -t http marian-mentoring https://www.marian.coach/mcp/mentoring
Claude.ai / Claude Desktop — Settings → Connectors → Add custom connector → paste https://www.marian.coach/mcp/mentoring
Cursor — add to .cursor/mcp.json:
{ "mcpServers": { "marian-mentoring": { "url": "https://www.marian.coach/mcp/mentoring" } } }
ChatGPT (developer mode) — Settings → Connectors → Add → MCP server URL https://www.marian.coach/mcp/mentoring
Microsoft 365 Copilot (via Copilot Studio) — open your agent → Tools → Add a tool → New tool → Model Context Protocol → Server URL https://www.marian.coach/mcp/mentoring, authentication None → Add to agent. Streamable HTTP, which is the one transport Copilot Studio supports.
Perplexity (Pro/Enterprise) — profile → All settings → Connectors → Custom connector → Remote → MCP Server URL https://www.marian.coach/mcp/mentoring, transport Streamable HTTP, authentication None.
No AI tool that supports MCP? The same wizard runs as a chat on marian.coach/mentoring-chat, and the classic pages live at /pricing/.
The read-only tools double as GET endpoints for scripts: https://www.marian.coach/mcp/mentoring/api/options, /api/match?role_band=em&motivation=scale-jump, /api/program?offer_id=first-quarter&start_date=2026-09-01, /api/verify (claim-code check). Spec: openapi.json. Sending an offer stays on the MCP tool and the chat — the doors that carry the discount.
Marian's general engineering-leadership toolkit (salary calculators, team-lead readiness, business-case builder — 9 tools) runs at marian.coach/mcp — same pattern, complementary tools.
Every price comes from the published mentoring catalog — the same generated file marian.coach renders, so this server cannot quote a price the site disagrees with. Track-record figures come from Marian's session log, pulled live on the site.
Built and maintained by Marian Kamenistak. Questions: marian@marian.coach