Humanity4AI
An open-source MCP (Message Context Protocol) server that equips AI agents with 9 humane skills — from supportive crisis replies to accessibility audits. Built to make every AI interaction safer, more inclusive, and more human.
MCP Endpoint
https://humanity4ai.ascent.partners/api/mcpProtocol: MCP 2024-11-05 · Server: humanity4ai v1.0.6 · Transport: HTTP POST (JSON-RPC 2.0)
9 Humanity Skills
Supportive Reply
Detects distress and generates supportive, non-clinical replies with escalation guidance.
Rewrite Depression-Sensitive Content
Audits or rewrites text to remove harmful language patterns around mental health.
Empathetic Reframe
Reframes messages with genuine empathy — acknowledging emotions and validating experience.
De-escalation Plan
Generates structured de-escalation guidance calibrated to conflict intensity.
Cultural Context Check
Checks content for cultural sensitivity issues for a given audience or region.
Cognitive Accessibility Audit
Audits content for reading level, structure, and clarity to reduce cognitive load.
Neurodiversity Design Check
Audits UI descriptions for ADHD, autism, dyslexia, and sensory awareness.
Age-Inclusive Design Check
Audits user flows and interfaces for age-inclusive design across all age groups.
Accessibility Audit (WCAG)
Crawls or audits pages against all 86 WCAG 2.2 success criteria.
Live Test Console
Send a JSON-RPC request to the MCP endpoint and see the response. Edit the payload below to explore available methods.
Response will appear here...