{
"markdown": "# Humanity4AI ⭐\n\n**9 humanity skills for AI agents** — crisis detection, accessibility auditing, empathy, cultural sensitivity, and more. Ready-to-use via MCP, npm, or direct prompting.\n\n[](https://github.com/humanity4ai/project_human/actions/workflows/ci.yml)\n[](https://www.npmjs.com/package/@humanity4ai/mcp-servers)\n[](LICENSE)\n[](https://github.com/humanity4ai/project_human/releases)\n[](https://github.com/humanity4ai/project_human?tab=readme-ov-file#quick-start)\n[](https://github.com/humanity4ai/project_human/actions/workflows/codeql.yml)\n[](https://registry.modelcontextprotocol.io?search=io.github.humanity4ai%2Fproject-human)\n\n\n\n---\n\n## Skills at a Glance\n\n| Skill | Category | What it does |\n|-------|----------|-------------|\n| 🛡️ **Supportive Reply** | Emotional Safety | Detects crisis signals, generates supportive responses with escalation guidance |\n| 📝 **Safe Content Rewriter** | Emotional Safety | Audits and rewrites text to remove harmful, stigmatising, or triggering patterns |\n| ♿ **Accessibility Audit** | Accessibility | Scores pages against all 86 WCAG 2.2 success criteria (A/AA/AAA) |\n| 🧠 **Cognitive Accessibility** | Cognitive Support | Audits content for reading level, structure, and cognitive load |\n| 🌍 **Cultural Context Check** | Cultural Context | Flags cultural sensitivity issues for a given audience and region |\n| 🔥 **De-escalation Plan** | Conflict Navigation | Generates structured de-escalation plans calibrated to conflict intensity |\n| 💬 **Empathetic Reframe** | Communication | Reframes messages with genuine empathy, catching hollow empathy patterns |\n| 🧩 **Neurodiversity Design** | Neurodiversity | Audits UIs for ADHD, autism, dyslexia, and sensory sensitivity |\n| 👶 **Age-Inclusive Design** | Age Inclusion | Audits user flows for age barriers across children, adults, and older users |\n\n> ⭐ **If you find this useful, a star helps others discover it**\n\n[](https://www.star-history.com/#humanity4ai/project_human&Date)\n\n---\n\n## Quick Start\n\n**Zero setup — just a URL:**\n\n```json\n{\n \"mcpServers\": {\n \"humanity4ai\": {\n \"url\": \"https://humanity4ai.ascent.partners/api/mcp\"\n }\n }\n}\n```\n\n**Or one command with npx:**\n\n```bash\nnpx @humanity4ai/mcp-servers\n```\n\nThen configure your MCP client:\n\n```json\n{\n \"mcpServers\": {\n \"humanity4ai\": {\n \"command\": \"npx\",\n \"args\": [\"-y\", \"@humanity4ai/mcp-servers\"]\n }\n }\n}\n```\n\n**Docker**: `docker compose up` \n**For contributors**: `git clone … && pnpm install && pnpm start` \n\nAll 9 skills are discoverable via `tools/list` and invocable via `tools/call`.\n\n**Prerequisites (for local only):** Node.js >= 22, pnpm >= 10 | Windows, macOS, Linux, Android, iOS\n\n---\n\n## Architecture\n\n```mermaid\ngraph TD\n A[MCP Client
VS Code / Cursor / Claude Code] -->|stdio| B[mcp-server.ts]\n A -->|HTTP POST| C[Vercel
humanity4ai.ascent.partners]\n B --> D[server-factory.ts
createServer]\n C -->|api/mcp.ts| D\n D --> E[9 Tool Registrations]\n E --> F[handlers.ts
invokeAction]\n F --> G[validate.ts
SCHEMA_REGISTRY]\n G --> H[schemas-data.ts
18 inline schemas]\n F --> I[9 Handler Functions
rule-based, zero LLM calls]\n I --> J[crisis-resources
patterns
i18n
wcag-criteria]\n```\n\n## Four Ways to Use\n\n| Method | Best for | How |\n|--------|----------|-----|\n| **Remote URL** | Zero-setup, any MCP client | Point to `https://humanity4ai.ascent.partners/api/mcp` |\n| **MCP Server** | VS Code, Cursor, Claude Code, Copilot, Manus AI, OpenCode | `npx @humanity4ai/mcp-servers` or clone & start |\n| **LLM Prompting** | ChatGPT, Claude, Gemini (web chat) | Share `llms.txt` or paste `llms-full.txt` into the chat |\n| **Local Files** | Offline CLI tools | Clone the repo, point your tool at the `skills/` directory |\n\nSee [Agent Adapter Guide](docs/agent-adapters.md) for platform-specific setup instructions.\n\n---\n\n## Why Humanity4AI?\n\nAI agents are everywhere — but they're not always humane. Humanity4AI gives agents **reusable, tested skills** for the moments that matter:\n\n- An agent detects a user in crisis → **Supportive Reply** generates an appropriate response and escalates to qualified help\n- A UI is inaccessible to screen readers → **Accessibility Audit** scores it against WCAG 2.2 and provides remediation\n- Content uses stigmatising language → **Safe Content Rewriter** flags and rewrites harmful patterns\n- A conflict is escalating → **De-escalation Plan** generates structured, non-coercive guidance\n\nEvery skill includes:\n- **Explicit safety boundaries** — what the skill can and cannot do\n- **Uncertainty disclosure** — confidence level stated upfront (low/medium/high)\n- **Evaluation gates** — automated baseline checks for quality\n\nThese skills are **rule-based and non-clinical** — they do not provide diagnosis, treatment, or professional medical/legal advice.\n\n---\n\n## Supported Platforms\n\nOpenCode · Claude Code · Microsoft Copilot · Manus AI · OpenClaw · ChatGPT · Claude · Gemini\n\n---\n\n## Contribute\n\n```bash\ngit clone https://github.com//project_human.git\ncd project_human\ngit checkout -b my-contribution\n# Copy the skill template if adding a new skill:\ncp -r templates/skill skills/my-skill-name\npnpm check && pnpm evals && pnpm test # Run all checks before PR\n```\n\nOpen a PR targeting `main`. Browse [good first issues](https://github.com/humanity4ai/project_human/issues?q=is%3Aopen+label%3A%22good+first+issue%22) or read the [Contributing Guide](CONTRIBUTING.md).\n\n| Error | Fix |\n|-------|-----|\n| `ERR_PNPM_OUTDATED_LOCKFILE` | Run `pnpm install`, then commit `pnpm-lock.yaml` |\n| `pnpm: command not found` | `npm install -g pnpm` |\n| `pnpm evals` fails | `EVAL_REPORT=1 pnpm evals` — see `evals/reports/latest.md` |\n\n---\n\n## Resources\n\n- [Contributing Guide](CONTRIBUTING.md)\n- [Agent Adapter Guide](docs/agent-adapters.md)\n- [Governance](docs/GOVERNANCE.md)\n- [Roadmap](docs/ROADMAP.md)\n- [Release Process](docs/release-process.md)\n- [Public Site](https://humanity4ai.github.io/project_human/)\n\n---\n\n**[Share on X](https://twitter.com/intent/tweet?text=Humanity4AI%20%E2%80%94%209%20humanity%20skills%20for%20AI%20agents%20%F0%9F%A4%9D%0A%0ACrisis%20detection%2C%20WCAG%20audits%2C%20empathy%2C%20cultural%20sensitivity%2C%20and%20more.%0A%0Ahttps%3A%2F%2Fgithub.com%2Fhumanity4ai%2Fproject_human)** · MIT License · Copyright © 2026 Ascent Partners Foundation\n",
"bytes": 7072,
"sha": "df54e93911b0d11d938a75b1f9ca7e61118f80a42d7c01700cb8517fd3234855",
"repo_slug": "humanity4ai/project_human",
"fonte": "repo",
"truncated": false,
"api": "https://agentalog.com/api/listings/mcp_io_github_humanity4ai_project_human_eebfe973/readme"
}