n8n Automated Jira Ticket Responses with GPT-4 and Pinecone

Scheduled trigger → query assigned Jira issues → retrieve from Pinecone knowledge base → AI generates reply → auto-post Jira comment

Who

Technical teams needing automated initial responses for assigned Jira tickets

You get

Assigned tickets automatically receive AI-generated comments based on historical knowledge base at set intervals

You need

n8n · jira · pinecone

Steps

1. Configure Jira credentials and schedule trigger

   Create Jira credentials in n8n (API token and email). Select the registered Jira account ID in the Jira Issue List node. Add a Schedule Trigger node and set the execution interval (mind API rate limits).

2. Configure Pinecone vector database

   Register Pinecone credentials in n8n (API key and environment). In the Pinecone Vector Store node, enter your index name (replace placeholder <Pinecone Indexを入力>). Ensure your knowledge base is already indexed in Pinecone.

3. Query assigned issues and extract content with comments

   Jira Issue List node queries assigned issues using JQL 'assignee = <accountId>'. Loop Over Items processes each issue. Code1 node extracts issue id, summary, and description. Jira Issue Detail node fetches latest comments. Code2 node concatenates comments into text and extracts updateAuthor.accountId. check accountId node filters out placeholder accountId (continues when not equal to '＜アカウントIDを入力してください＞').

   Execution prompt · Source text。Code1 node JS Code field

   Copy prompt

   return {
     'id': $input.first().json.id,
     'summary': $input.first().json.fields.summary,
     'description': $input.first().json.fields.description,
   }

   Execution prompt · Source text。Code2 node JS Code field

   Copy prompt

   let comments = '';
   let accountId = '';
   for (const item of $input.all()) {
     comments += '## ' + item.json.updated
     comments += '\n'
     comments += item.json.body.content[0].content[0].text
     comments += '\n'
     comments += '\n'
   
     accountId = item.json.updateAuthor.accountId
   }
   
   return {
     'comments': comments,
     'accountId': accountId
   };

4. Configure and run AI Agent for retrieval and response generation

   AI Agent node receives title and description from Code1 and comment history from Code2. OpenAI Chat Model sub-node is configured with API key; model options include gpt-4.1-mini, gpt-4.1, or gpt-4. Pinecone Vector Store sub-node provides 'jira情報' retrieval tool in retrieve-as-tool mode; Embeddings OpenAI provides 1536-dimension vectors. AI Agent generates response based on systemMessage role and user prompt.

   Execution prompt · Source text。AI Agent node Text / User Message field (promptType=define)

   Copy prompt

   =# タイトル
   {{ $('Code1').first().json.summary }}
   
   # 説明
   {{ $('Code1').first().json.description }}
   
   # コメント履歴
   {{ $json.comments }}

   Execution prompt · Source text。AI Agent node Options > System Message field

   Copy prompt

   あなたはプロジェクトの開発者です。タスクがアサインされました。過去の知見を利用して対応をしてください。outputの内容をそのままJiraのコメント欄に記載します。また関連する過去の知見はIDも合わせて回答してください。

5. Post comment and optionally reassign

   Jira Add Comment node posts the AI Agent output as a comment on the original issue. Jira Assign node reassigns the issue to the accountId extracted during comment retrieval (if different from placeholder). Loop Over Items then proceeds to the next issue.

You are setting up an n8n workflow to automatically generate AI replies for assigned Jira tickets using a Pinecone knowledge base. First confirm you have: Jira API token, Pinecone API key and index name, OpenAI API key, and a pre-indexed Pinecone knowledge base. Then import the workflow JSON or configure nodes manually following the steps.