Zenda Web Console Assistant
Overview
The Zenda WebConsole Assistant is a natural language operational assistant embedded within the Mozenda Web Console. Rather than requiring manual navigation or API scripts, Zenda executes backend operations using Model Context Protocol (MCP) tool definitions and Mozenda APIs. The assistant is generally available to all Web Console customers.
.png)
Key Features & Capabilities
Zenda covers all available Mozenda API operations. Key capabilities include:
Tool Discovery & Parameter Checking: Ask Zenda to list all executable tools or detail required/optional parameters for specific tools (e.g., publisher update or sequence step configuration).
Agent & Job Management: Run, save, stop, resume, or cancel agents and sequences. Automatically set up agent schedules (e.g., daily or monthly runs), query job stats, or fetch active job counts.
Collections & Sequences: Create, clear, or delete collections. Manage sequence steps, reorder execution steps, and set concurrent runs.
Audit Scripts: Generate, enable, disable, list, or delete audit scripts directly via natural language.
Account & Department Administration: Create or manage departments directly from parent/master accounts.
Session Navigation & Reset: Use the Up and Down arrow keys to cycle through recently entered prompts during an active session, or click the Reset Chat button to clear the current conversation view. Persistent history is preserved up to account storage limits.
Step-by-Step How-To Guide
1. Launching Zenda & Discovering Capabilities
a. Log in to the Mozenda Web Console.
b. Click the Zenda Assistant 'Z' icon located at the bottom right of the screen, next to the support chat bubble.
c. Type "What tools can you use?" to view a complete list of active operational tools.
.png)
2. Inspecting Tool Parameters
Before executing a task, ask Zenda for parameter requirements:
a. Type a query such as: "What are the required parameters for publisher update?" or "Show me parameters for creating an agent schedule."
b. Zenda checks Swagger definitions and outlines required fields (e.g., schedule frequency, dates, or trigger types).
.png)
3. Executing Operational Commands
-
Enter your operational prompt using specific IDs or names where applicable:
a. Scheduling: "Add a schedule to agent 1658 to run daily starting September 15."
b. Collection Management: "Create a collection named AITestCollection with fields Name and Surname" or "Empty collection ID 1001."
c. Sequence Control: "Set sequence step 2 to run concurrently."
d. Audit Scripts: "Create an audit script for agent 1234."
-
Review the human-readable confirmation message returned in the chat.
-
Navigate to the target Web Console section (e.g., Collections, Agents, or Sequences) to verify updated assets.
.png)
Best Practices & Safety Guidelines
⚠️ Important Precautions
Be Precise with IDs: Ensure you specify whether an operation target is an Agent, Job, or Sequence (e.g., cancelling an agent running inside a sequence requires specifying the sequence or job ID).
Administrative Caution: Commands like department creation or job cancellation execute immediately upon request. Double-check parameters before submitting destructive requests.
Session Memory: Use Up/Down arrows to recall previous commands in your current session. Note that hard-refreshing the browser clears the local prompt recall buffer.