Every Monday, we pull back the curtain on what’s shipping in Workato Enterprise MCP.
This Week: Support Knowledge Management, Now Agent-Ready
One new pre-built MCP server is live today — Zendesk Knowledge Base. We now have 27 production-grade MCP servers available for Workato Enterprise MCP customers.
What’s New
Zendesk Knowledge Base — 8 tools
Give agents the ability to search, retrieve, and navigate the Zendesk Guide knowledge base, create new draft articles, and update existing ones — closing the loop between ticket resolution and knowledge capture.
View Zendesk Knowledge Base documentation →
Why This Matters
Support organizations have a knowledge problem that isn’t really about knowledge — it’s about timing. The moment a support agent resolves a ticket is the moment they have the most complete, accurate understanding of the issue and its solution. It’s also the moment they’re least likely to document it. There’s another ticket in the queue. The knowledge walks out with the context.
The Zendesk Knowledge Base MCP Server closes that gap. An agent resolves a ticket, finds no existing article covering the issue, creates a draft article from the resolution notes, and flags it for review — without leaving their AI workflow. The workflow that makes knowledge bases actually stay current now takes seconds instead of getting skipped.
But the real story this week is what Zendesk Knowledge Base makes possible when combined with Zendesk Ticket Management — already live in the catalog — and the rest of the Workato MCP portfolio.
With Zendesk Ticket Management handling the operational layer (queue review, triage, SLA monitoring, customer replies) and Zendesk Knowledge Base handling the knowledge layer (article search, content creation, updates), agents have full support operations coverage in a single governed workflow. Find the ticket. Resolve it. Document it. One conversation.
The composability unlocks the workflows that matter most:
- Zendesk Knowledge Base + Zendesk Ticket Management — the complete support loop: surface an overdue ticket, investigate similar past tickets, find the relevant KB article, add a resolution comment, update the article if it’s outdated. Workato now covers both sides of Zendesk.
- Zendesk Knowledge Base + Slack — a support lead identifies a pattern in resolved tickets, drafts a new KB article from the resolution approach, and posts a summary to the team channel. Institutional knowledge captured at the moment it exists, distributed immediately.
- Zendesk Knowledge Base + Confluence — customer-facing KB article in Zendesk, internal engineering runbook in Confluence, both updated from one agent conversation. Support and engineering knowledge in sync.
- Zendesk Knowledge Base + Jira — a recurring ticket pattern surfaces a product bug. The agent creates a Jira issue from the ticket context, then drafts a KB article with a workaround for customers while the fix is in flight. Customer-facing and internal actions in one workflow.
Real-World Use Cases
Zendesk Knowledge Base: For Support Agents and Team Leads
- Knowledge gap identification: “Find any existing articles about MFA setup errors. Show me what we have.”
- Article review: “Pull the full content of that article. I need to check whether it covers the new SSO provider we added last month.”
- Knowledge capture: “There’s nothing covering this issue. Create a draft article in the Authentication section — title it ‘MFA Error: TOTP App Out of Sync’ and use the resolution we just applied.”
- Article maintenance: “Update that article — add the section about the new authenticator app workflow we rolled out in March.”
- Label management: “Add the ‘needs-review’ label to this article so it shows up in the content audit queue.”
- Structure navigation: “What sections are in the Identity & Access Management category? I need to know where to file a new article.”
- Paired with tickets: “Pull ticket 8821. How did we resolve it? Is there a KB article covering this? No? Create a draft.”
Zendesk Knowledge Base creates articles in draft state only — create_article always drafts; publishing requires a human reviewer in the Zendesk UI. This is intentional: AI-generated content goes through a review step before it reaches customers. Label updates replace the entire label set — the server reads current labels via get_article before updating to preserve existing tags. Article search uses Zendesk’s full-text search engine; newly created or updated articles may not appear in search results immediately due to platform indexing delays.
What Makes These Enterprise-Ready
- Verified User Access: All tool calls execute under the authenticated user’s Zendesk identity. Article creation, updates, and label changes are attributed to the acting agent. No shared service accounts, no attribution gaps.
- Draft-first creation pattern: create_article always creates in draft state. Publishing is a deliberate human action in the Zendesk UI. AI-generated content doesn’t reach customers without review.
- Read-before-write for labels: Label updates replace the full label set. The server retrieves current labels via get_article before modifying them — existing labels are never silently dropped.
- Permission inheritance: Agents can only access and modify content their Zendesk role grants them. The server doesn’t elevate privileges or expose restricted content.
- Managed infrastructure: Workato hosts, scales, and maintains the server. No Zendesk API versions to track, no rate limits to manage, no infrastructure to provision.
- Composable: Chains with Zendesk Ticket Management, Slack, Confluence, and Jira for end-to-end support and knowledge workflows.
- Audit trails: Every tool call logged with full context — who, what, when, and what was returned.
Get Started
Zendesk Knowledge Base is available now for Workato customers with Enterprise MCP.
Install time: Minutes
Custom development: None
Deployment complexity: One-click activation
View Zendesk Knowledge Base documentation →
The bottom line: Other vendors are still announcing MCP roadmaps. We’re shipping production-ready servers every Monday. See you next Monday