Help us improve
Share bugs, ideas, or general feedback.
From zapier
Checks Zapier MCP setup health with three modes: dashboard health check, audit for waste/duplicates, and systematic troubleshooting.
npx claudepluginhub zapier/zapier-mcp --plugin zapierHow this skill is triggered — by the user, by Claude, or both
Slash command
/zapier:zapier-statusThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Three modes for monitoring and maintaining a Zapier MCP setup. Determine the mode from context, or ask if unclear.
Walks through Zapier MCP setup, authentication, and mode detection. Use for onboarding, reconnecting, or exploring what Zapier can do across 9,000+ apps.
Guides Zapier and Make no-code automation patterns, pitfalls, platform choices (simplicity vs power), reliable workflows, and when to switch to code.
Sets up auth, provides reusable Python/Node.js MCP helper, enables tool discovery, and handles responses for connecting AI agents to Power Automate via FlowStudio.
Share bugs, ideas, or general feedback.
Three modes for monitoring and maintaining a Zapier MCP setup. Determine the mode from context, or ask if unclear.
First, detect the server mode: If list_enabled_zapier_actions is available, the user is on Agentic mode. Otherwise, the user is on Classic mode where each configured action is its own MCP tool (e.g., slack_send_channel_message, gmail_find_email).
Trigger: "check my tools", "zapier status", "is everything working?", or any general status inquiry.
A quick dashboard view of the current state.
Check the available tools:
list_enabled_zapier_actions to get the inventory of enabled actions.app_action_name. The built-in get_configuration_url tool is always present when the server is connected.If no Zapier tools are available: report the connection status and suggest running zapier-setup (Classic) or calling get_zapier_skill with name "zapier-mcp-onboarding" (Agentic).
If tools/actions are available, build a summary by grouping actions by app (Agentic: from the list_enabled_zapier_actions response; Classic: from each tool's description):
For each app, show:
Format as a dashboard:
Zapier MCP status
=================
Server: connected
Total actions: 14 across 5 apps
Slack 3 actions (2 read, 1 write) ✓ working
Gmail 3 actions (1 read, 2 write) ✓ working
Google Cal 2 actions (1 read, 1 write) ✓ working
Jira 3 actions (2 read, 1 write) ✓ working
Google Docs 2 actions (1 read, 1 write) ✓ working
If any actions appear broken (based on recent errors or auth issues), flag them.
End with: "Everything looks good." or "Found [N] issues. Want me to diagnose them?"
Trigger: "audit my setup", "clean up my tools", "find duplicates", "what should I remove?"
Find inefficiencies: duplicate actions, unused tools, conflicts with native MCP servers.
Get the full inventory:
list_enabled_zapier_actions.Check for duplicates within Zapier MCP:
slack_find_message and slack_search_messages)Check for conflicts with native MCP servers:
.cursor/mcp.json, claude_desktop_config.json, .mcp.json — depends on the client)Check for unused or low-value actions:
Show the audit report:
Audit results
=============
Duplicates: 1 found
- Slack: "find_message" and "search_messages" overlap. Recommend removing "search_messages".
Native conflicts: 1 found
- Slack: native Slack MCP also configured. Consider removing Zapier Slack actions.
Cleanup candidates: 2 found
- Google Sheets "delete_row": rarely needed, can re-add on demand
- Jira "add_attachment": niche action, add when needed
Recommended removals: 4 actions
Want me to show you how to clean these up?
disable_zapier_action to remove the recommended actions directly in chat.get_configuration_url and direct them to the web UI to remove the recommended actions. List exactly which ones to remove.Trigger: "what's broken?", "my tools aren't working", "debug my MCP", or when a specific tool call has failed.
Systematic troubleshooting with error pattern matching.
Gather context: Ask what went wrong, or use the error from the current conversation.
Run diagnostics in order:
a. Connection check: Try calling get_configuration_url or any available Zapier tool. If nothing works, the problem is server-level (auth, config, network).
b. Action check:
list_enabled_zapier_actions to see if the action is enabled. If not, use discover_zapier_actions to find it and enable_zapier_action to add it.c. Auth check: Try calling a read action for the affected app (Agentic: execute_zapier_read_action; Classic: the specific read tool). If it returns an auth error, the app connection needs re-authentication.
d. Parameter check: Review the failing call's parameters. Common issues:
Match against known error patterns:
| Symptom | Likely cause | Fix |
|---|---|---|
| All tools fail | Server auth expired | Re-authenticate at mcp.zapier.com |
| One app fails, others work | App-level auth expired | Re-connect that specific app |
| Tool not found / unavailable | Action not configured | Agentic: discover_zapier_actions + enable_zapier_action. Classic: direct user to get_configuration_url to add it |
| "invalid params" | Wrong fields or format | Check the tool's parameter schema |
| Results are empty but expected data exists | Search too narrow or wrong field | Broaden the search or check field names |
| Timeout on execute | Server overloaded or action is slow | Retry once, then report if persistent |
| "rate limit exceeded" | Too many calls | Space out requests, wait 30 seconds |
| Works in one project, fails in another | Project-level vs global config | Check both project-level and global MCP config for the client |
"Here's what I found:
list_enabled_zapier_actions. On Classic, this means inspecting the available tool names.