Eight tools, the whole ticket lifecycle
The MCP surface is deliberately small and complete — everything an agent needs to manage a ticket from report to resolved:
list_projects
The projects this key is allowed to see, nothing more.
get_project
Full detail on one project — workflow, members, settings.
list_tickets
List tickets with filters an agent can reason about.
get_ticket
One ticket, complete: description, links, history.
search_tickets
Free-text search across titles and ticket IDs.
create_ticket
File new tickets with category, severity and description.
update_ticket
Change status, severity and assignee as work progresses.
list_users
Resolve teammates for assignment.
Connect Claude in two minutes
Add Tiqzen to
.cursor/mcp.json,
Claude Desktop or
Claude Code with one block of config — your personal MCP key comes from Settings →
Security → MCP / AI access in the app, and the exact block is shown right
next to it:
{
"mcpServers": {
"tiqzen": {
"command": "npx",
"args": ["tsx", "tools/mcp-ticketing-bridge/src/index.ts"],
"env": {
"TICKETING_API_BASE": "https://api.tiqzen.com",
"TICKETING_MCP_API_KEY": "mcp_••••"
}
}
}
} Any client in, one governed backlog out
Every MCP key is scoped to its assigned projects — an agent with one client project's key cannot list, read or touch any other. Isolation is enforced at the API layer, not by prompt convention. That's what makes it safe to hand an autonomous agent the backlog.
Tiqzen
MCP server
5 tools · per-key scope
What teams run over MCP
Create tickets from your AI
Describe a bug to any assistant, get a filed, categorized ticket. How it works →
MCP access is free — on every plan
Unlike trackers that gate agent access behind enterprise tiers, Tiqzen includes the full MCP server in the Free plan. Create a workspace, mint a key, connect Claude — before your coffee cools. See pricing →
Give your agents a backlog they can actually work
Full MCP server on the Free plan. Create a workspace, mint a key, connect Claude — before your coffee cools.