Set up the Task Manager AI agent

Implement with AI

Send this guide to your coding agent.

Prerequisite: Install Port MCP

Open in ClaudeOpen in CodexOpen in Cursor

Open plan mode if your tool supports it; otherwise present the plan below filled in and wait for my approval. Implement this Port guide in my org via MCP:

https://docs.port.io/guides/all/setup-task-manager-ai-agent

Read the raw markdown version at https://docs.port.io/guides/all/setup-task-manager-ai-agent.md - it contains every tab and code block without page markup.

Goal: get the guide's core flow working end-to-end in my org; adapting it to fit my existing setup takes priority over matching the guide 1:1.

Plan:
1. Confirm MCP is connected, in the right org, with sufficient permissions.
2. If the guide offers alternative implementation paths (tabs), pick the one matching my installed integrations and tools, confirm it with me, and implement only that path.
3. Diff the guide's data model (blueprints, properties, relations, workflows, actions, agents, automations, integrations, webhook data sources, secrets) against mine.
4. Propose adaptations for gaps, reusing existing blueprints/relations over guide-named duplicates.
5. Flag what needs a UI click, credential, or secret from me, testing MCP capability empirically before ruling anything out. If the guide has a "Set up via API" section, use it for anything MCP can't do before treating a step as UI-only.
6. Stop on any blocker and give me options. Approving this plan authorizes the writes it lists; pause only for writes beyond what's listed.

Build:
- Extend blueprint schema additively when upserting; don't remove or overwrite existing properties, and treat type conflicts as a blocker, not an auto-fix.
- Never print secret values into the chat or logs; ask me to set them in Port, or write them via the secrets API without echoing them back.
- List any mock data in the plan, minimal and labeled mock; once approved, seed it without re-asking, and tell me what you seeded.
- For anything the guide writes downstream (e.g. a webhook target), use a real entity, not a mock.
- For pages/widgets, use the real page identifier from the app URL, not a guessed slug.
- When you hit a UI step confirmed (not assumed) unsupported via MCP and not covered by the guide's API sections, pause, give exact clicks, then resume via MCP.
- Validate and give links after each meaningful step (only a tool-returned URL, no guessed paths); don't proceed if the last run wasn't a success.

Done:
- Run the guide's "Let's test it" steps where possible (e.g. execute a workflow test run) and confirm the expected output exists in Port.
- Summarize adaptations, seeded data, what was mocked or skipped, remaining UI steps, and how to verify.

Overview

This guide will walk you through setting up a "Task Manager" AI agent in Port.
By the end of this guide, your developers will be able to get information about their tasks via Port's AI chat.

Common use cases

• Get a quick overview of assigned tasks and their priorities.
• Monitor pull requests waiting for review.
• Track task progress and status updates.

Set up data model

To create a Task Manager AI agent in Port, we'll need to configure two main components as described in our Build an AI agent guide:

• The data sources it will use to answer questions about tasks and pull requests.
• The agent configuration that defines its capabilities and conversation starters.

Configure data source access

For this guide, we will be using GitHub and Jira as our data sources to provide comprehensive task management capabilities. These integrations will automatically create and configure all the necessary resources needed by the Task Manager AI agent.

Install the following integrations to have access to these data sources:

• Port's GitHub Ocean integration for pull requests and code changes.
• Jira integration for task and issue tracking.

Optional tools

While this guide uses GitHub and Jira, you can choose tools that best fit your organization's needs. For example:

• GitLab or Azure DevOps instead of GitHub.
• Opsgenie instead of PagerDuty.

Create the agent configuration

1. Go to the AI Agents page in Port.

2. Click on + AI Agent.

3. Toggle Json mode on.

4. Copy and paste the following JSON schema:

   Task Manager agent configuration (Click to expand)

   {
     "identifier": "task_manager",
     "title": "Task Manager",
     "icon": "Details",
     "properties": {
      "description": "Task Manager responsible for answering questions about Jira issues, repositories (including READMEs), pull requests, services, and teams.",
      "status": "active",
      "prompt":"You are an agent responsible for answering questions about Jira issues, Pull Requests, Repositories, and teams.\n### Guidelines \n - Provide clear information about active issues (can be also referred as open issues ) with statuses: To Do and In Progress \n - Provide clear information about completed issues (can be also referred as close issues ) with statuses: Closed and Done \n - Identify pull requests that require attention (open pull requests)  \n - Provide clear information about repositories like their related service, README, team (in case \"owning team\" is empty, provide the \"Team\" related to the \"gitHub Teams\" property), last contributer, etc. \n - Identify connections between repositories, pull requests and issues to services\n\nWhen needed, you can create new Jira issues by calling the appropriate action.",
      "execution_mode": "Approval Required",
      "conversation_starters": [
        "Which tasks are assigned to me",
        "How many tasks are currently in progress",
        "Which PRs should I review?"
      ],
      "tools": [
        "^(list|search|track|describe)_.*",
        "^run__createJiraIssue$"
      ]
    }
   }

   MCP Enhanced Capabilities

   The AI agent uses MCP (Model Context Protocol) enhanced capabilities to automatically discover important and relevant blueprint entities via its tools. The ^(list|search|track|describe)_.* pattern allows the agent to access and analyze related entities in your context lake, such as Jira issues, GitHub pull requests, repositories, and teams. Additionally, we allow the corresponding self-service action in the agent's Allowed Actions input (or include ^run__createJiraIssue$ in the tools list in Advanced mode (regex)), which instructs the AI agent to call this specific action when needed to create new Jira issues.

5. Click on Create to save the agent.

Interact with the Task Manager

You can interact with the task manager AI agent in several ways.
This guide will demonstrate the two main ways.

Port UI

The Task Manager AI agent can be accessed through an AI Agent widget in your Port dashboard.
Follow the step below to set it up:

1. Go to the homepage in Port

2. Click on + Widget.

3. Choose AI agent.

4. Type Task Manager for Title.

5. Select Task Manager from the Agent dropdown.

   

6. Click on Save.

Once the widget is set up, you can:

• Use the conversation starter buttons to quickly check:

  • Your assigned tasks.
  • Work in progress.
  • Pull requests needing review.

• Type custom questions in the chat field about:

  • Tasks and tickets.
  • Pull requests.

• Engage in natural follow-up conversations to explore specific topics.

Slack Integration

You can also ask task-related questions through the Port Slack App. Before using this method, ensure you have installed and configured the app.

You can interact with Port in two ways:

1. Direct message the Port AI Assistant.
2. Mention the app in any channel it's invited to.

When you send a message, the app will:

1. Open a thread.
2. Respond with an answer.

Example queries:

@Port What tasks are assigned to me?
@Port Show me all PRs waiting for my review
@Port How many tasks are currently in progress?

Slack routes to the general AI assistant

The Slack app routes all requests to the Port AI general assistant, not to specific AI agents. The assistant can answer questions about your context lake, but it does not invoke the Task Manager agent or any other configured agent directly. For agent-specific interactions, use the AI Agent widget in the Port UI.

Example questions

Here are some questions you can ask the Task Manager agent:

• "Which of Jane Doe's pull requests need attention?".
• "Who are the code owners for the billing service repository?".
• "Which repository contains the code for the PaymentProcessor service?".
• "How many issues were closed by the frontend team in March 2025?".
• "What are the unresolved issues in my current sprint?".
• "Please open a Jira task to add a new payment method".

Best practices

To get the most out of your Task Manager agent:

1. Try it out: Start with simple queries and see how the agent responds.
2. Add context: If the response isn't what you expected, try asking again with more details.
3. Troubleshoot: If you're still not getting the right answers, check our troubleshooting guide for common issues and solutions.

Possible enhancements

You can further enhance the Task Manager setup by:

• Integration expansion: Add more data sources like GitLab or Azure DevOps.
• Automated notifications: Configure the agent to proactively notify about important updates.
• Custom conversation starters: Add organization-specific queries to the conversation starters.
• Monitor and improve: Check how your developers are interacting with the agent and improve it according to feedback.