MCP + REST API

The EQ layer for
ClaudeChatGPTCursorWindsurfSlackHubSpotevery AI model
every AI. any surface.

Qualia is the emotional intelligence API for AI — an EQ layer for AI agents that adds relationship intelligence and communication style awareness to any agent, assistant, or workflow. Understand people, not just prompts.

🤖 Claude
✨ ChatGPT
⌨️ Cursor
🌊 Windsurf
💬 Slack
📊 HubSpot
🧠 every AI model
Architecture

Two primitives.
Infinite surfaces.

Connect Qualia to Claude, ChatGPT, Cursor, Slack, or any MCP-compatible agent in minutes. The same relationship intelligence — wherever communication happens.

Every surface consumes the same two core calls: context.person to profile the recipient, and adapt.message to rewrite your draft.

🤖
✨
💬
✉️
📬
📊
⌨️
🌊
Q
The gap

Every AI can write.
None of them understand.

Today's agents are syntactically fluent and emotionally blind. They treat a frustrated parent, a skeptical VP, and an anxious student as the same input.

The missing layer isn't a better model. It's relationship intelligence — knowing who you're talking to, how they communicate, and what the relationship needs right now.

Before → After
WithoutGenerates a response to the message
With QualiaUnderstands who sent it before writing a word
WithoutSame tone for every recipient
With QualiaAdapts to each person's archetype and strategy
WithoutNo relationship history
With QualiaSurfaces risk signals and trust context
WithoutYou rewrite every draft
With QualiaDrafts you send with confidence
Core API Primitives — five calls. complete relationship intelligence.
POST
/v1/context/person
Infer a recipient's archetype, dimensions, and communication strategy from any message sample.
email + sample_text
archetype · strategy · dimensions
POST
/v1/adapt/message
Rewrite any draft to match the recipient's archetype, emotional state, and relationship context.
person_id · draft · intent
adapted_message · confidence · reasoning
GET
/v1/relationship/:id
Retrieve trust score, interaction history, and preferred channel for a relationship over time.
person_id
trust_score · history · channel
POST
/v1/conversation/analyze
Classify intent, detect reply necessity, and identify adjacency pairs in any message.
subject · body
intent_category · reply_necessity
POST
/v1/feedback
Submit accepted/edited/ignored signal to continuously improve future adaptations.
request_id · event · final
accepted
GET
/v1/health
Liveness check. Returns API version and uptime. No auth required.
—
status · version · uptime

Claude becomes
relationship-aware
in two lines.

Connect via our remote MCP server — Claude gains get_relationship_context and adapt_message as native tools. Zero prompt engineering required.

Or use the REST API directly. Python and JavaScript SDKs ship with typed responses, auto-retry, and the full OpenAPI 3.1 spec.

Or install the CLI: npx qualia profile sarah@acme.com and get a full archetype profile in your terminal in seconds.

# pip install qualia-sdk from qualia import Qualia client = Qualia(api_key="qua_...") ctx = client.context.person( email="sarah@acme.com", sample_text=thread.last_message ) # → { archetype: "Dragon", communication_strategy: [...] } draft = client.adapt.message( "Can you send the forecast by Friday?", recipient_id=ctx.person_id, intent="request" ) print(draft.adapted_message)
Use cases — how teams wire Qualia into their workflows
✉️
Email agents that feel human
/v1/context/person · /v1/adapt/message

An AI agent that writes emails without Qualia is syntactically correct and emotionally blind. With Qualia, the agent calls /v1/context/person before drafting — getting the recipient's archetype and communication strategy. It then calls /v1/adapt/message to rewrite the draft to fit.

ctx = client.context.person(email=<Str c="&quot;sarah@acme.com&quot;" />, sample_text=thread) draft = client.adapt.message(recipient_id=ctx.person_id, intent=<Str c="&quot;follow_up&quot;" />, message=draft_text)
💼
CRM copilots that know the relationship
/v1/context/person · /v1/relationship/state · /v1/adapt/message

Embed Qualia into a CRM side panel. When a rep opens a contact, the copilot surfaces relationship health, tone trend, and communication archetype — then suggests an adapted outreach draft.

state = client.relationship.state(person_id="per_9f2b1f7a") # → { health: 72, risk_signals: ["tone_shift"] } draft = client.adapt.message(recipient_id="per_9f2b1f7a", intent="re-engage", message=rep_draft)
🎧
Support agents that de-escalate
/v1/analyze/tone · /v1/adapt/message

Before a support agent replies to any ticket, Qualia runs /v1/analyze/tone — detecting frustration level, urgency, and churn risk. The agent then uses /v1/adapt/message to reframe the response so it lands with empathy, not just accuracy.

tone = client.analyze.tone(message=ticket_body) # → { tone: "frustrated", urgency: "high", risk: "churn_signal" } if tone.urgency == "high": flag_for_human_review(ticket_id)
🤖
Claude tasks that understand who they're talking to
get_relationship_context (MCP) · adapt_message (MCP)

Wire Qualia to any Claude workflow via the remote MCP server. Claude gains get_relationship_context and adapt_message as native tools — no prompt engineering required. Zero EQ logic for the developer.

# Claude API call — Qualia MCP connected # Claude automatically calls: # get_relationship_context(email="marcus@acme.com") # adapt_message(recipient_id=..., message=draft) # No extra code required.
🎓
Teacher and counselor communication at district scale
/v1/context/person · /v1/adapt/message · /v1/briefing

A teacher writing to a frustrated parent, a counselor following up with an anxious student — all high-stakes communications that happen dozens of times a day. The SDK enables LMS and SIS platforms to embed this district-wide.

ctx = client.context.person(email=parent_email, sample_text=parent_message) draft = client.adapt.message( recipient_id=ctx.person_id, intent="academic_update", message=teacher_draft, context="parent_communication" )
The 12 Qualia archetypes
D
Dragon
Charismatic Leader

Bottom line first. Action items second. Details optional.

P
Phoenix
Strategic Executor

Decision. Reasoning. Implementation.

G
Griffin
Protective Leader

Trust first. Execution second.

L
Leviathan
Commanding Influencer

Future-state focused. Big opportunities.

R
Raven
Relationship Builder

Use "we." Build consensus.

O
Oracle
Insightful Communicator

Context first. Reasoning second. Conclusion third.

S
Sentinel
Dependable Protector

Provide plan. Provide ownership.

W
Wolf
Community Builder

Focus on people. Show support.

C
Centaur
Disciplined Guide

Provide framework. Provide process. Provide next steps.

S
Sphinx
Analytical Strategist

Data. Analysis. Recommendation.

S
Sage
Intellectual Explorer

Teach. Explain. Explore.

O
Owl
Precision Keeper

Provide details. Provide documentation. Avoid ambiguity.

Where it runs — any surface. any agent. any platform.
🤖
Claude
MCP live
✉️
Gmail
Extension live
📬
Outlook
Add-in live
✨
ChatGPT
Q3
⌨️
Cursor
Q3
💬
Slack
Q3 pilot
🟦
Teams
Q3 pilot
📊
HubSpot
Q3
☁️
Salesforce
Q4
📁
Google Docs
Q3
🌊
Windsurf
Q3
🎓
LMS / SIS
Roadmap

Start building in
under 10 minutes.

Free sandbox. No credit card. Python + JS SDKs. Claude MCP recipe on day one.