Understand the workspace
List projects and issues, inspect scans and pages, read reports and VPAT® records, search documentation, and look up WCAG criteria or account limits.
New · Remote MCP access
Accessibility MCP server for real remediation work
Connect an MCP-compatible AI agent to your Accessibility Tracker workspace. Review WCAG issues, compare scans, plan fixes, and move work forward without separating the agent from your audit evidence or project permissions.
- Read-only by default
- Revocable scoped keys
- Included on paid plans
AI agent
Connected to Accessibility Tracker
Compare the last two scans and build a remediation plan for regressions.
14 new findings across 6 pages since the previous scan.
3 regressions also map to open audit issues in the project backlog.
Remediation plan grouped by owner, user impact, and WCAG criterion.
Project access enforcedSource: scan + audit backlog
More Than Another Scan Command
Browser testing MCP servers are useful for finding machine-detectable problems on a page. Tracker connects the agent to the longer remediation lifecycle around them.
A scan-only workflow
- Finds machine-detectable problems on the current page.
- Requires project history and ownership to be supplied separately.
- Stops before validation, reporting, and documentation.
Accessibility Tracker MCP
- Retrieves audit issues, scan history, owners, comments, and validation status.
- Compares results over time and connects regressions to existing work.
- Supports remediation plans, reports, VPAT® workflows, and controlled actions.
Agent-ready context
Tools for Understanding, Planning, and Acting
Focused tools, reusable resources, and guided prompts let the agent request structured project data instead of scraping the dashboard.
Plan with evidence
Use guided prompts to triage a scan, prepare a remediation plan, summarize regressions, or draft VPAT® remarks for human review.
Act when permitted
Optional capabilities support creating and updating issues, adding comments, and importing scan findings, and starting and cancelling scan jobs. You choose them when you create a key, and permanent deletion is never exposed.
Connect With Deliberate Control
Start with the smallest useful permission set and expand it only when the workflow requires more.
Create a scoped key
Generate a key in Settings. Read access is always included; write and scan-job capabilities are explicit choices when available.
Connect your MCP client
Add the remote HTTP endpoint and keep the key in an environment variable instead of pasting it into prompts or source files.
Work from real project context
Ask about issues, scans, pages, reports, VPAT® records, documentation, WCAG criteria, or account limits.
Approve deliberate actions
When enabled, the agent can create or update issues, add comments, import findings, and manage scan jobs without bypassing your project access.
The endpoint
Point any MCP client that speaks streamable HTTP at this URL and send your key as a bearer token. Settings gives you copy-paste configuration for Claude Code, Cursor, VS Code and Windsurf once you have a key.
POST https://accessibilitytracker.com/api/mcp
Authorization: Bearer <your key>
Content-Type: application/jsonThe MCP server documentation has the full tool reference, error codes, rate limits, and a curl handshake you can run before configuring a client.
Built for Constrained Agent Access
MCP does not create a shortcut around Tracker permissions. Every request is authenticated, scoped to the connected account, and checked against normal project authorization.
- Read-only access on every key; writes and scan jobs are separately granted.
- Keys are shown once, stored as hashes, and can be revoked from Settings.
- No hard-delete tools are exposed through MCP.
- Rate limits, idempotency controls, and untrusted-content fencing reduce agent risk.
- Your MCP client remains responsible for asking you to approve sensitive tool calls.
Included, not an add-on
MCP Access Starts at $19 per Month
The Small Business plan and every higher tier include the Accessibility Tracker MCP server. The Free tier is a good way to explore Tracker, but it does not include MCP keys.
Questions
Accessibility MCP Server Questions
What to know before connecting an AI agent to Tracker.
What is an accessibility MCP server?
An accessibility MCP server exposes accessibility data and workflows to an AI client through the Model Context Protocol. Accessibility Tracker uses it to give an agent structured access to your projects, WCAG issues, scans, reports, and documentation instead of making the agent rely on copied-and-pasted context.
Does the MCP server replace manual accessibility audits?
No. Automated scans and AI assistance cannot determine full WCAG conformance. Tracker keeps audit evidence and human validation at the center, while MCP helps your team retrieve context, plan remediation, and complete controlled workflow steps faster.
Which plans include MCP access?
MCP access is included with the Small Business, Mid-Sized Company, Enterprise, and Enterprise Mega plans. It is not included in the Free tier.
Can an AI agent change my accessibility data?
Not by default. New keys begin with read access. Write and scan-job capabilities must be selected for the key and enabled by the service before those tools are available. Your MCP client should still ask you to approve tool calls that change data.
How is access secured?
Each connection uses a revocable API key tied to your account. Requests keep normal project and tenant authorization, capabilities are separated by key, rate limits apply, and destructive hard-delete tools are not exposed.
Give Your Agent Better Accessibility Context
Start with Tracker, choose a paid plan, and connect from Settings when MCP is enabled for your workspace.