Surface resource actions from Jira with AI

Developers and platform engineers often spend significant time requesting cloud resources and permissions needed to complete a task. That involves context-switching, manual lookup, and filing platform requests, which slows delivery.

This guide shows how to use Port's AI agent to analyze a Jira issue, infer required resources and permissions, and create AI Suggested Action entities that surface the right self-service actions (e.g., create S3 bucket, grant IAM permissions). Teams can then run or reject suggested actions from the catalog and use a dashboard to review pending approvals.

Common use cases

• Analyze Jira issue with AI agent – Trigger the AI agent to process new Jira tickets and generate suggested actions.
• Run or reject suggested actions – Review, execute, or reject platform actions directly from the AI Suggested Action entity.
• Monitor suggested actions dashboard – Track pending approvals, action statuses, and trends from the review board dashboard.

Prerequisites

This guide assumes the following:

• You have a Port account and have completed the onboarding process.
• You have access to create and configure AI agents in Port.
• Jira issues are synced into Port (e.g., via the Jira integration) and you have a jiraIssue blueprint.
• You have self-service actions in Port for the resources and permissions you want to suggest (e.g., create S3 bucket, grant IAM permissions).

Set up data model

You will create the AI Suggested Action blueprint to store each suggestion from the AI agent, linked to the source Jira issue.

Create the AI Suggested Action blueprint

1. Go to the builder page of your portal.

2. Click on + Blueprint.

3. Click on {...} Edit JSON.

4. Use the following configuration:

   AI Suggested Action blueprint (click to expand)

   {
     "identifier": "aiSuggestedAction",
     "description": "This blueprint represents an AI suggested action",
     "title": "AI Suggested Action",
     "icon": "AI",
     "schema": {
       "properties": {
         "actionIdentifier": {
           "title": "Action Identifier",
           "type": "string"
         },
         "actionProperties": {
           "title": "Action Properties",
           "type": "object"
         },
         "entityIdentifier": {
           "title": "Entity Identifier",
           "type": "string"
         },
         "confidence": {
           "title": "Confidence",
           "type": "number"
         },
         "reasoning": {
           "title": "AI Reasoning",
           "type": "string",
           "format": "markdown"
         },
         "status": {
           "icon": "DefaultProperty",
           "title": "Status",
           "type": "string",
           "enum": [
             "suggested",
             "executed",
             "rejected"
           ],
           "enumColors": {
             "suggested": "yellow",
             "executed": "green",
             "rejected": "red"
           }
         }
       },
       "required": []
     },
     "mirrorProperties": {},
     "calculationProperties": {},
     "aggregationProperties": {},
     "relations": {
       "jiraIssue": {
         "title": "Jira Issue",
         "target": "jiraIssue",
         "required": true,
         "many": false
       }
     }
   }

5. Click Create to save the blueprint.

Set up self-service actions

You will create three self-service actions: one for the AI agent to create suggested action records (AI-facing), one for users to run the suggested platform action (human-facing), and one for users to reject a suggestion (human-facing).

Create AI Suggested Action

1. Go to the Self-service page of your portal.

2. Click on + Action.

3. Click on the {...} Edit JSON button.

4. Use the following configuration:

   Create AI Suggested Action action (click to expand)

   {
     "identifier": "create_ai_suggested_action",
     "title": "Create AI Suggested Action",
     "icon": "Actions",
     "description": "Creates a new AI suggested action record",
     "trigger": {
       "type": "self-service",
       "operation": "CREATE",
       "userInputs": {
         "properties": {
           "jira_issue": {
             "type": "string",
             "title": "Jira Issue ID",
             "description": "The identifier of the Jira issue"
           },
           "action_identifier": {
             "icon": "DefaultProperty",
             "type": "string",
             "title": "Action Identifier",
             "description": "The Identifier of the suggested self service action in Port"
           },
           "action_properties": {
             "icon": "DefaultProperty",
             "type": "object",
             "title": "Action Properties",
             "description": "The action's input properties"
           },
           "entity_identifier": {
             "type": "string",
             "title": "Entity Identifier",
             "description": "Entity identifier if the action is a DAY-2"
           },
           "confidence": {
             "type": "number",
             "title": "Confidence",
             "description": "Confidence level of the suggested action"
           },
           "ai_reasoning": {
             "type": "string",
             "title": "AI Reasoning",
             "description": "Reasoning behind the suggested action",
             "format": "markdown"
           }
         },
         "required": [
           "action_identifier",
           "action_properties"
         ],
         "order": [
           "action_identifier",
           "action_properties",
           "entity_identifier",
           "confidence",
           "ai_reasoning"
         ]
       }
     },
     "invocationMethod": {
       "type": "UPSERT_ENTITY",
       "blueprintIdentifier": "aiSuggestedAction",
       "mapping": {
         "identifier": "{{.inputs.jira_issue}}-{{.inputs.action_identifier}}",
         "title": "{{.inputs.jira_issue}}-{{.inputs.action_identifier}}",
         "properties": {
           "actionIdentifier": "{{.inputs.action_identifier}}",
           "actionProperties": "{{.inputs.action_properties}}",
           "confidence": "{{.inputs.confidence}}",
           "entityIdentifier": "{{.inputs.entity_identifier}}",
           "reasoning": "{{.inputs.ai_reasoning}}",
           "status": "suggested"
         },
         "relations": {
           "jiraIssue": "{{.inputs.jira_issue}}"
         }
       }
     },
     "requiredApproval": false,
     "allowAnyoneToViewRuns": true
   }

5. Click Create to save the action.

Run Suggested Platform Action

1. Go to the Self-service page of your portal.

2. Click on + Action.

3. Click on the {...} Edit JSON button.

4. Use the following configuration:

   Run Suggested Platform Action action (click to expand)

   {
     "identifier": "run_suggested_platform_action",
     "title": "Run Suggested Platform Action",
     "trigger": {
       "type": "self-service",
       "operation": "DAY-2",
       "userInputs": {
         "properties": {},
         "required": [],
         "order": []
       },
       "blueprintIdentifier": "aiSuggestedAction"
     },
     "invocationMethod": {
       "type": "WEBHOOK",
       "url": "https://api.port.io/v1/actions/{{.entity.properties.actionIdentifier}}/runs",
       "agent": false,
       "synchronized": true,
       "method": "POST",
       "headers": {
         "RUN_ID": "{{ .run.id }}",
         "Content-Type": "application/json"
       },
       "body": {
         "properties": "{{.entity.properties.actionProperties}}"
       }
     },
     "requiredApproval": false,
     "allowAnyoneToViewRuns": true,
     "description": "Execute the platform action recommended by the AI agent for this Jira issue.",
     "icon": "Cloud"
   }

5. Click Create to save the action.

Reject AI Suggested Action

1. Go to the Self-service page of your portal.

2. Click on + Action.

3. Click on the {...} Edit JSON button.

4. Use the following configuration:

   Reject AI Suggested Action action (click to expand)

   {
     "identifier": "reject_ai_suggested_action",
     "title": "Reject AI Suggested Action",
     "description": "Reject the suggested platform action generated by the AI agent.",
     "trigger": {
       "type": "self-service",
       "operation": "DAY-2",
       "userInputs": {
         "properties": {},
         "required": [],
         "order": []
       },
       "blueprintIdentifier": "aiSuggestedAction"
     },
     "invocationMethod": {
       "type": "UPSERT_ENTITY",
       "blueprintIdentifier": "aiSuggestedAction",
       "mapping": {
         "identifier": "{{.entity.identifier}}",
         "properties": {
           "status": "rejected"
         }
       }
     },
     "requiredApproval": false,
     "allowAnyoneToViewRuns": true,
     "icon": "Unavailable"
   }

5. Click Create to save the action.

Configure the AI agent

You will create the Resource Suggestion Agent that analyzes Jira issues, discovers available self-service actions via list_actions, and creates one AI Suggested Action entity per suggested action by calling create_ai_suggested_action.

1. Go to the AI agents page of your portal.

2. Click on + AI Agent.

3. Click on the {...} Edit JSON button.

4. Use the following configuration:

   Resource Suggestion Agent (click to expand)

   {
     "identifier": "resource_suggestion_agent",
     "title": "Resource Suggestion Agent",
     "icon": "Bolt",
     "team": [],
     "properties": {
       "description": "AI agent that analyzes Jira-linked work items and suggests relevant Port self-service actions for cloud resources and permissions",
       "status": "active",
       "prompt": "You are a Resource Suggestion Agent for work items linked to Jira issues.\n\nYour job is to analyze Jira issues and determine which self-service actions should be suggested in order to complete the task described in the issue.\n\nUse the `list_actions` MCP tool to discover available self-service actions in the portal. Match inferred needs to available actions and suggest which actions the user should run.\n\nEach suggested action must be saved as a separate entity in the \"aiSuggestedAction\" blueprint by ALWAYS calling the `create_ai_suggested_action`\n\nFor every suggested action you must provide:\n\n1. action_identifier: The identifier of the Port self-service action.\n\n2. action_properties: A JSON object containing the input properties required by that action.\n\n3. entity_identifier (optional): Only include this field if the action is a DAY-2 operation that targets an existing entity (for example scaling an EC2 instance or modifying a service).\n\n4. confidence: A number between 0 and 100 representing how confident you are that this action is correct.\n\n5. ai_reasoning: A short explanation describing why the action is needed based on the Jira issue content.\n\n6. jira_issue: The linked Jira issue identifier\n\nRules:\n\n* Suggest only actions that clearly map to the problem described in the Jira issue.\n* If the issue requests a new infrastructure resource, do NOT include entity_identifier.\n* If the issue modifies or operates on an existing resource, include entity_identifier.\n* If multiple resources are required, create multiple suggested actions.\n* If no platform action is needed, don't call the `create_ai_suggested_action`",
       "execution_mode": "Automatic",
       "conversation_starters": [
         "What resources and permissions does this ticket need?",
         "Which self-service actions should I run for this Jira issue?",
         "Analyze this ticket and surface relevant cloud resource actions."
       ],
       "tools": [
         "^(list|get|search|track|describe)_.*",
         "run_create_ai_suggested_action"
       ]
     },
     "relations": {}
   }

5. Click Create to save the agent.

How the agent creates suggestions

The agent uses the list_actions MCP tool (available via Port's AI context) to discover self-service actions in your portal. For each action it recommends, it calls create_ai_suggested_action with the action identifier, input properties, optional entity identifier for DAY-2 actions, confidence, reasoning, and Jira issue ID. Each call creates or updates one AI Suggested Action entity linked to the Jira issue.

Set up automation

You will add two automations: one to invoke the AI agent when a Jira issue is created, and one to mark the AI Suggested Action as executed when the user runs the suggested platform action.

Trigger when Jira issues are created

1. Go to the Automations page of your portal.

2. Click on + Automation.

3. Click on {...} Edit JSON.

4. Use the following configuration. You can add a condition with JQ expressions to control when the agent is invoked (e.g., only for certain projects or issue types).

   On Jira Issue Created automation (click to expand)

   {
     "identifier": "jira_issue_discover_actions",
     "title": "On Jira Issue Created",
     "description": "Automation that triggers AI agent to auto discover and provision cloud resources in a Jira ticket",
     "icon": "AI",
     "trigger": {
       "type": "automation",
       "event": {
         "type": "ENTITY_CREATED",
         "blueprintIdentifier": "jiraIssue"
       },
       "condition": {
         "type": "JQ",
         "expressions": [],
         "combinator": "and"
       }
     },
     "invocationMethod": {
       "type": "WEBHOOK",
       "url": "https://api.getport.io/v1/agent/resource_suggestion_agent/invoke",
       "agent": false,
       "synchronized": true,
       "method": "POST",
       "headers": {
         "RUN_ID": "{{ .run.id }}",
         "Content-Type": "application/json"
       },
       "body": {
         "prompt": "Here is the Jira issue information:\n\nJira Issue ID: {{.event.diff.after.identifier}}\nJira Issue Title: {{.event.diff.after.title}}\n\nRelated Entities in the Catalog (for context): {{.event.diff.after.relations}}\n\nFull Jira Issue Properties:\n{{.event.diff.after.properties}}",
         "labels": {
           "source": "auto_discover_actions",
           "jira_issue_id": "{{.event.diff.after.identifier}}"
         }
       }
     },
     "publish": true,
     "allowAnyoneToViewRuns": true
   }

5. Click Create to save the automation.

Mark executed suggested action

1. Go to the Automations page of your portal.

2. Click on + Automation.

3. Click on {...} Edit JSON.

4. Use the following configuration:

   Update AI Suggested Action Status automation (click to expand)

   {
     "identifier": "update_ai_suggested_action_status",
     "title": "Update AI Suggested Action Status",
     "description": "Automation to set the status of the AI suggested action after the platform engineer executes the action",
     "trigger": {
       "type": "automation",
       "event": {
         "type": "RUN_UPDATED",
         "actionIdentifier": "run_suggested_platform_action"
       },
       "condition": {
         "type": "JQ",
         "expressions": [],
         "combinator": "and"
       }
     },
     "invocationMethod": {
       "type": "UPSERT_ENTITY",
       "blueprintIdentifier": "aiSuggestedAction",
       "mapping": {
         "identifier": "{{ .event.diff.after.entity.identifier }}",
         "properties": {
           "status": "executed"
         }
       }
     },
     "publish": true,
     "allowAnyoneToViewRuns": true
   }

5. Click Create to save the automation.

Set up dashboard

With your data model, actions, and automations in place, you can create a dedicated dashboard to review and manage AI-generated platform action suggestions from Jira issues.

Create "Suggested action review board" dashboard

1. Navigate to the Catalog page of your portal.
2. Click on the + button in the left sidebar.
3. Select New dashboard.
4. Name the dashboard Suggested Action Review Board.
5. Input Review and manage AI-generated platform action suggestions derived from Jira issues under Description.
6. Select the Cloud icon.
7. Click Create.

You now have a blank dashboard where you can add widgets to visualize suggested actions and pending approvals.

Add widgets

In the new dashboard, create the following widgets:

Dashboard description (click to expand)

1. Click + Widget and select Markdown.

2. Copy and paste the following content:

   ## What this dashboard shows
   
   This dashboard gives your team visibility into AI-suggested platform actions from Jira issues.
   
   - Review suggested actions (create S3 bucket, grant IAM permissions, etc.) and run or reject them from one place.
   - Track pending approvals and see which actions are most often suggested.
   
   Built with Port to reduce platform request friction and speed up access to cloud resources.

3. Click Save.

Total actions pending approval (click to expand)

1. Click + Widget and select Number Chart.
2. Title: Total actions pending approval (add the Cloud icon).
3. Select Count Entities as the Chart type and choose AI Suggested Action as the Blueprint.
4. Add a filter: Property = status, Operator = equals, Value = suggested.
5. Click Save.

Suggested action by status (click to expand)

1. Click + Widget and select Pie Chart.
2. Title: Suggested action by status (add the AI icon).
3. Choose AI Suggested Action as the Blueprint.
4. Select Status as the Property to aggregate by (suggested, executed, rejected).
5. Click Save.

Most suggested platform actions (click to expand)

1. Click + Widget and select Bar Chart.
2. Title: Most suggested platform actions (add the Bolt icon).
3. Choose AI Suggested Action as the Blueprint.
4. Select Action Identifier (or the property that holds the Port action identifier) as the Property to aggregate by.
5. Click Save.

Pending approvals table (click to expand)

1. Click + Widget and select Table.
2. Title: Pending approvals (add the Checklist icon).
3. Choose the AI Suggested Action blueprint.
4. Add a filter: Property = status, Operator = equals, Value = suggested.
5. Click Save to add the widget to the dashboard.
6. Click on the ... button in the top right corner of the table and select Customize table.
7. In the top right corner of the table, click on Manage Properties and add the following properties:
   • Action Identifier: The Port self-service action suggested.
   • AI Reasoning: Why the action was suggested.
   • Confidence: The AI confidence score.
   • Jira Issue: The linked Jira issue.
   • Status: Current status (suggested, executed, rejected).
8. Click on the save icon in the top right corner of the widget to save the customized table.

Test the workflow

1. Create or sync a Jira issue into Port that describes a need for a cloud resource or permission (e.g., "Need an S3 bucket for feature X").
2. If the automation is enabled, the Resource Suggestion Agent runs automatically, otherwise invoke the agent manually with the Jira issue context.
3. Confirm that AI Suggested Action entities are created and linked to the Jira issue, with actionIdentifier, actionProperties, reasoning, and status = suggested.
4. Open an AI Suggested Action entity and run Run Suggested Platform Action or Reject AI Suggested Action.
5. After running the suggested action, verify the automation updates the entity status to executed (or confirm rejected after using Reject).
6. Open the Suggested Action Review Board dashboard and confirm the widgets show the expected counts and pending approvals.
7. Use the AI invocations page to debug agent runs, tool calls, and failures.

Related guides

• Sync Jira to your software catalog.
• AI-powered work item preparation.
• Triage tickets to coding agents.