BetterUp MCP needs a coaching action-permission map

Inventory actions by participant effect

List every current and proposed MCP tool with its exact client, account, input, data read, output, external effect, system of record, user confirmation, administrator control, and failure mode. Separate retrieving a resource, asking for coaching support, viewing availability, booking or changing a session, messaging a human coach, and writing a goal or progress record. Similar interface labels can create very different obligations for the participant and organization.

For each action, state whether the user may inspect, draft, confirm, cancel, correct, and appeal it. Identify actions unavailable in the tested client and those described only as future work. Pin client and server versions and reopen the map when tools, schemas, prompts, write capability, authentication, record location, or supported clients change.

Bind authority to the person and purpose

A successful login proves an account relationship, not permission for every coaching action. Record who initiated the connection, whether participation is voluntary, the program purpose, employer and provider roles, administrator restrictions, data categories, intended recipients, retention, access, and a simple disconnect path. Make clear what an employer, manager, administrator, coach, AI system, and participant can and cannot see or change.

Keep coaching support distinct from performance management, employment decisions, health care, therapy, crisis response, and legal advice. Define escalation and human-contact paths for sensitive, urgent, ambiguous, or out-of-scope situations. Do not infer a concern, score a person, or route an employment consequence from tool activity unless a separately authorized and appropriate process supports it.

Test the complete client-to-coaching path

Run representative reads and actions through the exact supported client and account type. Verify authentication discovery, token scope and expiry, resource binding, consent and confirmation screens, tool selection, arguments, duplicate and retry behavior, time zones, cancellations, provider readback, audit events, error messages, revocation, and what remains after disconnect. Test a denied action and a mistaken booking as carefully as a successful request.

A stateless server can still participate in a stateful experience because the client, identity provider, coaching platform, calendar, message store, and logs may retain data. Diagram those systems and confirm current contractual and technical handling. Limit high-sensitivity text in prompts and tool arguments, use the minimum scope, and keep real secrets outside content and model context.

Evaluate coaching value separately from connectivity

Measure successful and failed connections, unauthorized or unexpected tool calls, booking changes, duplicate actions, user corrections, human handoffs, support cases, access and accessibility, participant experience, coaching engagement, and the bounded development outcome the program was designed to support. Connection volume and resource retrieval describe activity. They do not prove coaching quality, behavior change, manager effectiveness, or business value.

Compare the MCP workflow with the provider's ordinary app, calendar, messaging path, human coordinator, and no-integration option. Expand only if the connection reduces participant friction without weakening consent, confidentiality, control, or coaching continuity. BetterUp's article documents provider capabilities and plans; buyer tests and participant evidence determine fit.

Keep roadmap capability outside present approval

Maintain separate rows for live, pilot, announced, planned, and unknown tools. A planned ChatGPT or Microsoft Copilot connection, richer prompt, deeper booking flow, or future goal write should not inherit approval from current Claude reads. Require a new source review, contract and configuration check, representative test, participant notice, permission decision, and rollback for every material expansion.

The program owner, privacy and security teams, coaching lead, HR or learning owner, accessibility lead, procurement, legal counsel, and participant representatives should review the parts they own. Record unresolved questions and restrict the connection until evidence arrives. A broad MCP approval is too coarse for a domain where the same endpoint can move from reading resources to changing a person's coaching record.

Turn this source into a reviewable decision

For AI Coaching Platforms for Leadership Development, use this briefing as a dated decision record rather than a substitute for the source. Preserve BetterUp: Introducing the BetterUp MCP server, the exact URL, the October 8, 2026 review date, the supported facts above, the editorial interpretation, the limitations, and any buyer-specific evidence. Link that record to the decisions most directly affected: Workflow and identity integration; Data flow and confidentiality; Safety, boundaries, and escalation; Accessibility, language, culture, and user control. State whether the source changes the scope, evidence requirement, control, sequence, or only the language used to describe the decision.

Before action, name the accountable owner, affected population and workflow, exact offering or configuration, source data and rights, human decision point, exception and appeal path, complete cost, expected benefit, failure and stop conditions, retained evidence, and next review date. Keep official facts, provider statements, buyer observations, representative tests, measured outcomes, editorial inferences, and unknowns visibly separate. Reopen the record when the source, offer, model, integration, data, policy, population, responsible person, or measured result changes.

Limitations and unknowns

BetterUp is the provider source. Its October 9, 2025 article describes its MCP server, current Claude support, setup paths, named tool examples, provider security architecture statements, and future plans. The page may be updated as clients and workflows change. It does not establish a buyer's entitlement, administrator configuration, exact live tool inventory, participant consent, token and client behavior, data flows across every connected service, confidentiality, retention, access, scheduling accuracy, message delivery, future write availability, coaching quality, human handoff, employment appropriateness, behavior change, or organizational outcome. Current contracts and documentation, exact client and server configuration, tool schemas, participant-facing notices, representative action and revocation tests, audit and support evidence, and qualified coaching, learning, HR, accessibility, privacy, security, identity, procurement, records, and legal review control.