About Axcess
Axcess is a paid MCP server for designers and developers evaluating typography accessibility. It assesses design-related text criteria, including font-weight-aware contrast, spacing, line length, and hierarchy, and returns a structured report.
Key features
- 01Evaluate 1–50 typography elements
- 02Assess contrast with font-weight context
- 03Check text size and line height
- 04Review line length, casing, and spacing
- 05Assess text on varied backgrounds
- 06Return structured accessibility reports
Use cases
- 01Review typography in a screen design
- 02Identify accessibility issues missed by automated scanners
MCP support
Axcess exposes `list_capabilities` and `evaluate_typography` tools to MCP clients. Typography evaluations cost $0.05 USDC via x402 on Base mainnet.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
Tools
3Saved tool list · updated 3w ago
This tool takes no parameters.
Details
From the project READMEAxcess — Typography Accessibility Evaluation via MCP
A paid MCP server that evaluates typography for accessibility issues beyond what automated scanners catch.
$0.05 USDC per evaluation · Base mainnet · x402 micropayments
What It Catches
Automated tools (axe, Lighthouse, WAVE) catch roughly 30% of WCAG failures and almost none of the design-judgment failures. Axcess evaluates:
- Contrast ratio with font-weight context — thin fonts require higher ratios than bold fonts at the same size
- Minimum font size for body text, captions, and labels
- Line height adequacy (including for dyslexic users)
- Line length (measure) comfort — too wide or too narrow
- Extended all-caps and italic use
- Letter spacing extremes
- Text on gradient, image, or pattern backgrounds
- Heading hierarchy and visual distinction between levels
- Display type scaling and tracking
- Bringhurst and Lupton design-craft rules beyond WCAG
Connect as an MCP Client
Claude.ai (web — Pro / Max / Team / Enterprise)
Settings → Integrations → Add MCP Server
https://axcess-mcp-server.fly.dev/mcp
Claude Desktop / Cursor (via mcp-remote)
Add to claude_desktop_config.json (or Cursor MCP settings):
{
"mcpServers": {
"axcess": {
"command": "npx",
"args": [
"mcp-remote",
"https://axcess-mcp-server.fly.dev/mcp"
]
}
}
}
Programmatic (AI agents and scripts)
Use @x402/fetch to handle payment automatically:
import { wrapFetchWithPaymentFromConfig } from '@x402/fetch';
import { ExactEvmScheme } from '@x402/evm/exact/client';
import { privateKeyToAccount } from 'viem/accounts';
const account = privateKeyToAccount(process.env.EVM_PRIVATE_KEY as `0x${string}`);
const fetchWithPayment = wrapFetchWithPaymentFromConfig(fetch, {
schemes: [{ network: 'eip155:*', client: new ExactEvmScheme(account) }],
});
const response = await fetchWithPayment(
'https://axcess-mcp-server.fly.dev/mcp',
{
method: 'POST',
headers: {
'content-type': 'application/json',
'accept': 'application/json',
'mcp-protocol-version': '2025-03-26',
},
body: JSON.stringify({
jsonrpc: '2.0',
id: 1,
method: 'tools/call',
params: {
name: 'evaluate_typography',
arguments: {
screen_name: 'Hero Section',
elements: [
{
element_type: 'heading',
heading_level: 1,
font_size: 48,
font_weight: 300,
line_height: 1.1,
color_hex: '#888888',
background_color_hex: '#ffffff',
},
],
},
},
}),
}
);
See src/scripts/test_paid_tool.ts for a complete working example.
Pricing
| Tool | Price | Status |
|---|---|---|
list_capabilities | Free | Live |
evaluate_typography | $0.05 USDC | Live |
evaluate_accessibility | $0.10 USDC | Coming soon |
Payment: USDC on Base mainnet (eip155:8453) via x402 protocol.
Tools
list_capabilities (free)
Returns available tools, pricing, and rubric categories. No payment required.
evaluate_typography ($0.05)
Evaluates 1–50 typography elements and returns a structured accessibility report.
Input:
{
screen_name?: string; // optional label for the report
elements: Array<{
element_id?: string;
element_type: 'heading' | 'body' | 'caption' | 'label' | 'button' | 'display' | 'ui';
heading_level?: 1 | 2 | 3 | 4 | 5 | 6;
font_size: number; // px
font_weight: number; // 100–900
line_height: number; // multiplier (e.g. 1.5, not px)
letter_spacing?: number; // em units
text_transform?: 'none' | 'uppercase' | 'lowercase' | 'capitalize';
font_style?: 'normal' | 'italic' | 'oblique';
content_length?: number; // character count
color_hex: string; // '#rrggbb'
background_color_hex: string;
background_type?: 'solid' | 'gradient' | 'image' | 'pattern' | 'video';
context?: string; // e.g. 'hero section', 'navigation'
}>;
}
Output:
{
screen_name?: string;
evaluated_at: string;
element_evaluations: Array<{
element_id: string;
element_type: string;
issues: Array<{
criterion_id: string;
criterion: string;
severity: 'critical' | 'major' | 'minor';
wcag_reference: string;
what_scanners_miss: string;
finding: string;
recommendation: string;
passes_wcag_minimum: boolean;
passes_design_judgment: boolean;
}>;
score: number; // 0–100
passes: string[];
}>;
summary: {
overall_score: number;
verdict: 'pass' | 'needs_work' | 'fail';
total_issues: number;
by_severity: { critical: number; major: number; minor: number };
};
top_issues: Issue[];
}
Deploy to Fly.io (free tier)
Requirements: flyctl, USDC wallet on Base, free CDP API key
fly auth login
fly launch --name axcess-mcp-server --no-deploy
# Set secrets (never committed to git)
fly secrets set PAYMENT_ADDRESS=0x...
fly secrets set X402_NETWORK=eip155:8453
fly secrets set CDP_API_KEY_ID=...
fly secrets set CDP_API_KEY_SECRET=...
fly deploy
Your app will be live at https://axcess-mcp-server.fly.dev.
Update the URL in smithery.yaml and the Connect section above before submitting to registries.
Self-Hosting Locally
Requirements: Node.js 18+
git clone <repo>
cd axcess-mcp-server
cp .env.example .env
# Edit .env with your PAYMENT_ADDRESS, CDP keys, and X402_NETWORK
npm install
npm run build
npm start
Local dev on testnet (no real payments):
PAYMENT_ADDRESS=0x... X402_NETWORK=eip155:84532 npm run dev
EVM_PRIVATE_KEY=0x... npm run test:paid
See .env.example for all configuration options.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.