compare_profiles
ChatGPTGet VITALS' structured profile-to-profile comparison for one connection returned by list_profile_comparisons. Use this for questions about how the pair differs, aligns, communicates, collaborates, or encounters friction. This returns VITALS' own calibrated conclusions and interpretation—not raw profile scores. Person A is always the user who authorized the call, and person B is the other person — every A-side field, from the facet positions to the narrative's standing instructions, belongs to the caller. Never invent a comparison when the status is not ready. On an acquaintance pair the narrative is null by design and the cards over the facet map are the whole read; answer from them and keep claims on the traits.
get_360_feedback
ChatGPTGet this user's VITALS 360, where people they invited placed them on the same scales the user answered about themselves — the same results they see in the VITALS app. Use it for anything about how others actually experience them, or whether their self-read matches the outside read. It answers at any stage: how many more people must answer, whether a round is ready for the user to close, and the full results once they unlock. Pass round with a title from a previous result to read one round; omit it for the most recent round whose results have unlocked. Rounds are never combined. Reviewers answer anonymously, so individual answers and rater identity are never returned — what reviewers wrote reaches you only through the generated cards. Ground every claim about what reviewers said in what this returns, and never guess at results that have not unlocked.
get_career_profile
ChatGPTGet this user's VITALS Career Profile: how they operate at work, the conditions they thrive in, what sustains and what drains them over time, and how their communication lands in common professional situations — the same view they see in the VITALS app. Use it for anything about their working life: role fit, team dynamics, managing up, burnout, interview and performance conversations. Prefer it over their general personality profile for work questions, and note it does not cover specific career paths or job recommendations. A status of 'not_generated' means they have not built it yet in the VITALS app — say so rather than guessing at it; 'no_profile' means they have not completed their VITALS assessment. Nothing here is saved back: if they tell you something new about their work, tell them to add it in VITALS.
get_collaboration_directives
ChatGPTGet up to five concrete, imperative instructions for how to work with this specific user, derived from their most distinctive VITALS traits. These are behavioral cues for you, the assistant — apply them to tune how you communicate and collaborate (e.g. how direct to be, how much detail to give, whether to lead or defer).
get_more_tools
ChatGPTCheck for additional tools whenever your task might benefit from specialized capabilities - even if existing tools could work as a fallback.
get_personality_profile
ChatGPTRetrieve the connected user's VITALS personality profile: their scored and banded traits across six dimensions (Values, Interests, Temperament, Action Style, Learning Style, Social Style), their headline identity statement, cross-cutting patterns, and per-dimension summaries — the same view they see in the VITALS app. Use it to personalize tone, framing, and approach. Bands are tendencies, never strengths or weaknesses.
get_relationship_read
ChatGPTGet everything VITALS holds on one person the user tracks: the read it wrote for them (the situation, how the user's own tendencies collide with it, and concrete experiments to try), every fact the user has given, the questions still unanswered, and the experiments they chose to keep. Get the id from list_relationships. Speak from this read rather than re-deriving from their general profile. A status of 'pending' means the read is still being written — say so, never invent one; 'failed' means it did not come through, and they can retry it in the VITALS app. Nothing here is saved back: if the user tells you something new about this person, tell them to add it in VITALS.
get_relationship_tendencies
ChatGPTGet how this user tends to show up in close relationships generally, independent of any one person: where they sit on six relationship dynamics (what they do, how it lands on others, and what flips it), the combinations that make them distinctive, and how well-meant moves from other people tend to land on them. Use it to shape advice about handling conflict, closeness, or anyone in their life — their general personality profile does not cover any of this. It names nobody. A status of "unavailable" means it has not been built yet — say so rather than guessing at it.
list_profile_comparisons
ChatGPTList the people whose VITALS profiles this user is authorized to compare, with each comparison's relationship, its readiness, and the connection_id required by compare_profiles. Call this before compare_profiles unless a connection_id is already known.
list_relationships
ChatGPTList the people this user tracks in VITALS: each one's id, what they call them, the relationship type, their own description of who that person is, and whether a read has been written for them. Call this whenever the conversation turns to a specific person in the user's life, so you know whether they are tracked and what VITALS already holds. It returns no detail about any one person — use get_relationship_read for that.
ping
ChatGPTLiveness check for the VITALS MCP server. Returns pong.
