klbr/n8nworkflows.xyz
1
0
Fork 0
mirror of https://github.com/khoaliber/n8nworkflows.xyz.git synced 2026-04-28 00:29:22 +00:00
Code Issues Packages Projects Releases Wiki Activity

creation

Browse Source
This commit is contained in:
nusquama
2025-11-12 14:24:58 +01:00
parent daef625b42
commit ffd52cac53
1 changed files with 597 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
workflows/Automate Agile Refinement Prep with Gmail, OpenAI & Google Sheets-3909/readme-3909.md
+597
View File
@@ -0,0 +1,597 @@
Automate Agile Refinement Prep with Gmail, OpenAI & Google Sheets
https://n8nworkflows.xyz/workflows/automate-agile-refinement-prep-with-gmail--openai---google-sheets-3909
# Automate Agile Refinement Prep with Gmail, OpenAI & Google Sheets
### 1. Workflow Overview
This workflow automates the preparation for Agile backlog refinement sessions using Google Sheets, Gmail, OpenAI, and Google Calendar. Targeted at Scrum Masters, Agile Coaches, and Product Owners, it ensures high-quality, validated user stories are ready ahead of refinement meetings, streamlining collaboration and communication.
The workflow consists of the following logical blocks:
- **1.1 Trigger and Initialization:** Manual or scheduled start, environment setup.
- **1.2 Calendar Event Check:** Query Google Calendar for upcoming refinement events.
- **1.3 Backlog Retrieval and Filtering:** Extract user stories from Google Sheets backlog, filter by readiness and priority.
- **1.4 Definition of Ready (DoR) Validation:** Validate stories against DoR criteria using AI agents.
- **1.5 Multi-Perspective AI Feedback:** Obtain Scrum Master, Business, and Technical feedback on each story from OpenAI.
- **1.6 Feedback Aggregation and Backlog Update:** Merge AI feedback and update the backlog in Google Sheets.
- **1.7 Email Composition and Approval Flow:** Generate a structured HTML email summarizing refinement content, send for Scrum Master approval, and then either send or create a draft email.
- **1.8 Error Handling:** Capture failures and send error notifications via email.
---
### 2. Block-by-Block Analysis
#### 1.1 Trigger and Initialization
**Overview:**
Starts the workflow manually or via a scheduled trigger and sets up environment variables required for the entire flow.
**Nodes Involved:**
- When clicking ‘Test workflow’
- Schedule Trigger (disabled)
- Start Here: Set Environment Variables
- No Operation, do nothing
**Node Details:**
- **When clicking ‘Test workflow’**
- Type: Manual trigger node
- Role: Allows manual start for testing or ad-hoc runs
- Inputs: None
- Outputs: Connects to environment variable setup
- Edge Cases: None significant
- **Schedule Trigger**
- Type: Schedule trigger node (disabled)
- Role: Intended for scheduled runs (e.g., periodic checks)
- Inputs: None
- Outputs: Connects to environment variable setup if enabled
- Edge Cases: Disabled, no effect unless enabled and configured
- **Start Here: Set Environment Variables**
- Type: Set node
- Role: Defines workflow-wide variables such as sheet names, statuses, and other parameters
- Inputs: Trigger nodes
- Outputs: Connects to No Operation and calendar loading
- Edge Cases: Misconfiguration could break downstream logic
- **No Operation, do nothing**
- Type: NoOp node
- Role: Placeholder allowing parallel path to calendar loading
- Inputs: Environment setup node
- Outputs: Connects to Load Calendar of Scrum Master node
- Edge Cases: None
---
#### 1.2 Calendar Event Check
**Overview:**
Checks the Scrum Master's Google Calendar for upcoming refinement events to trigger the backlog processing only if relevant events exist.
**Nodes Involved:**
- Load Calendar of Scrum Master
- Check for Refinement Project A (Filter)
- If no event do nothing (NoOp)
- Take DoR link for further referral (Google Drive)
- Take backlog link for further referral (Google Drive)
**Node Details:**
- **Load Calendar of Scrum Master**
- Type: Google Calendar node
- Role: Fetches calendar events for the Scrum Master to detect refinement meetings
- Configuration: Uses OAuth2 credentials for Google Calendar; queries for events with specific naming conventions
- Inputs: No Operation node
- Outputs: Filter node to check for refinement event
- Edge Cases: API rate limits, auth failures, no events found
- **Check for Refinement Project A**
- Type: Filter node
- Role: Filters calendar events to find a refinement session for a specific project
- Inputs: Calendar events node
- Outputs: If event found, continues workflow; else triggers no-op
- Edge Cases: No matching event leads to no further processing
- **If no event do nothing**
- Type: NoOp node
- Role: Terminates workflow path if no refinement event is found
- Inputs: Filter node negative output
- Outputs: Connects to DoR link retrieval for possible logging or reference
- Edge Cases: None
- **Take DoR link for further referral**
- Type: Google Drive node
- Role: Retrieves the link to the Definition of Ready document for later reference
- Inputs: No Operation node after no event or calendar check
- Outputs: Passes link to backlog link retrieval node
- Edge Cases: Access permissions, file not found
- **Take backlog link for further referral**
- Type: Google Drive node
- Role: Retrieves the backlog document link for further processing
- Inputs: DoR link retrieval node
- Outputs: Connects to Google Sheets nodes fetching backlog stories
- Edge Cases: Permissions, file availability
---
#### 1.3 Backlog Retrieval and Filtering
**Overview:**
Pulls user stories from the Google Sheets backlog and filters them by their readiness and priority status for refinement consideration.
**Nodes Involved:**
- Select Stories Ready for Refinement from Backlog
- Select Stories High Prio To Do from Backlog
- Join Potential Stories for Refinement (Merge)
**Node Details:**
- **Select Stories Ready for Refinement from Backlog**
- Type: Google Sheets node
- Role: Reads user stories marked as ready for refinement based on configured criteria (e.g., status column)
- Configuration: Uses credentials with access to backlog sheet, queries rows filtered by status
- Inputs: Backlog link retrieval node
- Outputs: Merge node for combining with high priority stories
- Edge Cases: Sheet structure change, empty result sets
- **Select Stories High Prio To Do from Backlog**
- Type: Google Sheets node
- Role: Reads user stories marked as high priority and to-do, ensuring critical items are included
- Inputs: Backlog link retrieval node
- Outputs: Merge node
- Edge Cases: Same as above
- **Join Potential Stories for Refinement**
- Type: Merge node
- Role: Combines the two sets of stories (ready and high priority) into a single list without duplicates
- Inputs: Two Google Sheets nodes
- Outputs: Loop batches for DoR checking and Scrum feedback addition
- Edge Cases: Duplicate detection logic, large data sets
---
#### 1.4 Definition of Ready (DoR) Validation
**Overview:**
Validates each candidate user story against the Definition of Ready criteria using AI agents, ensuring stories meet the quality bar before refinement.
**Nodes Involved:**
- Loop Over Items for DoR check (SplitInBatches)
- Read DoR criteria (Google Sheets)
- OpenAI Scrum Master DoR Check (LM Chat OpenAI)
- Compare User Story to DoR Criterium (Langchain Agent)
- Aggregate DoR check to User Story Level (Aggregate)
- Add DoR feedback to User Story (Merge)
- Provide Scrum Master Feedback on the story (Langchain Agent)
- Update Scrum feedback in Backlog (Google Sheets)
**Node Details:**
- **Loop Over Items for DoR check**
- Type: SplitInBatches
- Role: Processes user stories in batches to optimize API calls and resource use
- Inputs: Merged stories node
- Outputs: Parallel outputs to DoR criteria reading and Scrum feedback addition
- Edge Cases: Batch size misconfiguration leading to timeouts
- **Read DoR criteria**
- Type: Google Sheets node
- Role: Retrieves the list of DoR criteria from a dedicated sheet for validation
- Inputs: Loop Over Items node (start)
- Outputs: Connects to AI agent for comparison
- Edge Cases: Criteria sheet changes or missing data
- **OpenAI Scrum Master DoR Check**
- Type: LM Chat OpenAI node
- Role: Uses OpenAI to evaluate if the user story meets each DoR criterium
- Configuration: Uses OpenAI credentials, custom prompt for DoR check
- Inputs: DoR criteria and user story data
- Outputs: Connects to comparison node
- Edge Cases: API errors, prompt misconfiguration
- **Compare User Story to DoR Criterium**
- Type: Langchain Agent node
- Role: Aggregates AI responses to determine compliance per story and criterium
- Inputs: AI DoR check node
- Outputs: Aggregation node for story-level summary
- Edge Cases: Data mismatch, aggregation errors
- **Aggregate DoR check to User Story Level**
- Type: Aggregate node
- Role: Combines criterium-level checks into a consolidated story readiness rating
- Inputs: Comparison node
- Outputs: Merge node for adding DoR feedback
- Edge Cases: Aggregation logic errors
- **Add DoR feedback to User Story**
- Type: Merge node
- Role: Attaches DoR feedback to the respective user story record
- Inputs: Aggregate node and Loop Over Items for Scrum feedback
- Outputs: Scrum Master feedback AI node
- Edge Cases: Data sync issues
- **Provide Scrum Master Feedback on the story**
- Type: Langchain Agent node
- Role: Generates detailed Scrum Master perspective feedback on the story and DoR status
- Inputs: Merged story with DoR feedback
- Outputs: Google Sheets update node
- Edge Cases: API failures
- **Update Scrum feedback in Backlog**
- Type: Google Sheets node
- Role: Writes back Scrum Master feedback into the backlog sheet to keep it up to date
- Inputs: Scrum Master feedback AI node
- Outputs: Loop Over Items for DoR check (cycle continuation)
- Edge Cases: Write permission, concurrent editing
---
#### 1.5 Multi-Perspective AI Feedback
**Overview:**
Collects business and technical validation from OpenAI agents on each user story to enrich refinement preparation with diverse expert feedback.
**Nodes Involved:**
- Loop Over Items Business Feedback (SplitInBatches)
- OpenAI Business Analyst (LM Chat OpenAI)
- Business Validation (Langchain Agent)
- Update Business Validation in Backlog (Google Sheets)
- Loop Over Items Technical Feedback (SplitInBatches)
- OpenAI Technical Analyst (LM Chat OpenAI)
- Development Team Validation (Langchain Agent)
- Update Technical Validation in Backlog (Google Sheets)
- Add Scrum Feedback to Story (Merge)
- Add Business feedback to Story (Merge)
- Add Technical feedback to the story (Merge)
- Combine stories into one list (Aggregate)
**Node Details:**
- **Loop Over Items Business Feedback**
- Type: SplitInBatches
- Role: Processes stories batch-wise for business feedback generation
- Inputs: Scrum feedback merge node
- Outputs: Business AI node and merge node
- Edge Cases: Batch sizing, API throttling
- **OpenAI Business Analyst**
- Type: LM Chat OpenAI
- Role: Generates business-related feedback on user stories
- Inputs: Business feedback loop
- Outputs: Business Validation agent node
- Edge Cases: API limits, prompt accuracy
- **Business Validation**
- Type: Langchain Agent
- Role: Processes AI output for business validation insights
- Inputs: Business Analyst node
- Outputs: Updates backlog sheet node
- Edge Cases: Parsing errors
- **Update Business Validation in Backlog**
- Type: Google Sheets node
- Role: Writes business feedback back to the backlog sheet
- Inputs: Business Validation agent
- Outputs: Loop Over Items Business Feedback continuation
- Edge Cases: Write conflicts
- **Loop Over Items Technical Feedback**
- Type: SplitInBatches
- Role: Processes stories batch-wise for technical feedback generation
- Inputs: Business feedback merge node
- Outputs: Technical AI node and merge node
- Edge Cases: Same as business feedback
- **OpenAI Technical Analyst**
- Type: LM Chat OpenAI
- Role: Provides technical feedback on user stories
- Inputs: Technical feedback loop
- Outputs: Development Team Validation agent node
- Edge Cases: API errors
- **Development Team Validation**
- Type: Langchain Agent
- Role: Processes technical AI responses
- Inputs: Technical Analyst node
- Outputs: Google Sheets update node
- Edge Cases: Data format issues
- **Update Technical Validation in Backlog**
- Type: Google Sheets node
- Role: Writes technical feedback back to backlog
- Inputs: Development Team Validation
- Outputs: Loop Over Items Technical Feedback continuation
- Edge Cases: Write permission issues
- **Add Scrum Feedback to Story**
- Type: Merge node
- Role: Joins Scrum feedback with the story data for further processing
- Inputs: Batch loops and story data
- Outputs: Business feedback merge node and loop continuation
- Edge Cases: Data mismatch
- **Add Business feedback to Story**
- Type: Merge node
- Role: Adds business feedback to story records
- Inputs: Business feedback loop and merge
- Outputs: Technical feedback loop node
- Edge Cases: Synchronization
- **Add Technical feedback to the story**
- Type: Merge node
- Role: Adds technical feedback to stories
- Inputs: Technical feedback loop and merge nodes
- Outputs: Aggregate node to combine all feedback
- Edge Cases: Data consistency
- **Combine stories into one list**
- Type: Aggregate node
- Role: Combines all feedback-enriched stories into a single collection for email drafting
- Inputs: Technical feedback merge node
- Outputs: Draft email AI node
- Edge Cases: Large data volume management
---
#### 1.6 Feedback Aggregation and Backlog Update
**Overview:**
After gathering all feedback, this block updates the backlog with all AI validations, ensuring the source of truth reflects the latest insights.
**Nodes Involved:**
(Described above in feedback update nodes)
- Update Scrum feedback in Backlog
- Update Business Validation in Backlog
- Update Technical Validation in Backlog
**Node Details:**
As described in previous block; each Google Sheets update node writes respective AI feedback back into the backlog to preserve data continuity and transparency.
---
#### 1.7 Email Composition and Approval Flow
**Overview:**
Generates a formatted email summarizing the refined backlog, requests Scrum Master approval, and upon approval, either sends the email or creates a draft for manual adjustment.
**Nodes Involved:**
- Draft Email for Refinement (Langchain Agent)
- OpenAI Scrum Master Emailer (LM Chat OpenAI)
- Ask Approval to Scrum Master (Gmail)
- If Email content is Approved (If)
- Send Email to Attendees (Gmail)
- Create Draft in Scrum Master Email (Gmail)
- Make aware the draft is ready to be adjusted manually (Gmail)
**Node Details:**
- **Draft Email for Refinement**
- Type: Langchain Agent
- Role: Drafts a structured, HTML-rich email summarizing stories and feedback for the refinement session
- Inputs: Aggregated story list with feedback
- Outputs: Gmail node requesting approval
- Edge Cases: Prompt clarity affecting email quality
- **OpenAI Scrum Master Emailer**
- Type: LM Chat OpenAI
- Role: Provides the text content or enhancements for the email draft
- Inputs: Aggregated stories
- Outputs: Draft Email node
- Edge Cases: API errors
- **Ask Approval to Scrum Master**
- Type: Gmail node
- Role: Sends email to Scrum Master requesting content approval; linked to webhook for response
- Inputs: Draft Email node
- Outputs: If node for approval decision
- Edge Cases: Gmail API limits, webhook failures
- **If Email content is Approved**
- Type: If node
- Role: Branches workflow based on Scrum Master's approval response
- Inputs: Approval Gmail node webhook
- Outputs: Sends email or creates draft
- **Send Email to Attendees**
- Type: Gmail node
- Role: Sends finalized email to team members for refinement session
- Inputs: If node approval branch
- Edge Cases: Email sending failures, rate limits
- **Create Draft in Scrum Master Email**
- Type: Gmail node
- Role: Creates an email draft in Scrum Master's mailbox for manual review/editing
- Inputs: If node rejection branch
- Outputs: Notification node
- **Make aware the draft is ready to be adjusted manually**
- Type: Gmail node
- Role: Sends notification email to Scrum Master indicating the draft is ready for manual adjustment
- Inputs: Draft creation node
- Edge Cases: Email delivery issues
---
#### 1.8 Error Handling
**Overview:**
Monitors workflow errors and sends notification emails to administrators or operators for quick response.
**Nodes Involved:**
- Error Trigger
- Send Error Email (Gmail)
**Node Details:**
- **Error Trigger**
- Type: Error Trigger node
- Role: Captures any unhandled errors in workflow execution
- Outputs: Sends error details to email node
- Edge Cases: None
- **Send Error Email**
- Type: Gmail node
- Role: Sends detailed error report to predefined recipients
- Configuration: Uses Gmail credentials; retries enabled
- Edge Cases: Email sending failures
---
### 3. Summary Table
| Node Name | Node Type | Functional Role | Input Node(s) | Output Node(s) | Sticky Note |
|----------------------------------|---------------------------------|-------------------------------------------------|-----------------------------------|---------------------------------------------------|-------------------------------------------|
| When clicking ‘Test workflow’ | Manual Trigger | Manual start trigger | None | Start Here: Set Environment Variables | |
| Schedule Trigger | Schedule Trigger (disabled) | Scheduled start trigger (disabled) | None | Start Here: Set Environment Variables | |
| Start Here: Set Environment Variables | Set | Define environment variables | Manual/Scheduled Trigger | No Operation, do nothing | |
| No Operation, do nothing | NoOp | Placeholder for parallel execution | Set Environment Variables | Load Calendar of Scrum Master | |
| Load Calendar of Scrum Master | Google Calendar | Fetch Scrum Master's calendar events | No Operation | Check for Refinement Project A | |
| Check for Refinement Project A | Filter | Filter events for refinement meeting | Load Calendar | If no event do nothing, or continue workflow | |
| If no event do nothing | NoOp | End path if no refinement event found | Filter | Take DoR link for further referral | |
| Take DoR link for further referral| Google Drive | Retrieve DoR criteria document link | NoOp | Take backlog link for further referral | |
| Take backlog link for further referral| Google Drive | Retrieve backlog document link | DoR link retrieval | Select Stories Ready for Refinement from Backlog, Select Stories High Prio To Do from Backlog | |
| Select Stories Ready for Refinement from Backlog | Google Sheets | Extract stories ready for refinement | Backlog link retrieval | Join Potential Stories for Refinement | |
| Select Stories High Prio To Do from Backlog | Google Sheets | Extract high priority to-do stories | Backlog link retrieval | Join Potential Stories for Refinement | |
| Join Potential Stories for Refinement | Merge | Combine story lists without duplicates | Two Google Sheets nodes | Loop Over Items for DoR check, Add Scrum Feedback to Story | |
| Loop Over Items for DoR check | SplitInBatches | Batch processing of stories for DoR check | Merge node | Read DoR criteria, Add Scrum Feedback to Story | |
| Read DoR criteria | Google Sheets | Retrieve DoR criteria from sheet | Loop Over Items for DoR check | Compare User Story to DoR Criterium | |
| OpenAI Scrum Master DoR Check | LM Chat OpenAI | AI evaluation of user story against DoR criteria | Read DoR criteria | Compare User Story to DoR Criterium | |
| Compare User Story to DoR Criterium | Langchain Agent | Aggregate AI DoR evaluations | OpenAI DoR Check | Aggregate DoR check to User Story Level | |
| Aggregate DoR check to User Story Level | Aggregate | Consolidate DoR results per user story | Comparison node | Add DoR feedback to User Story | |
| Add DoR feedback to User Story | Merge | Attach DoR feedback to user story | Aggregate node, Loop Over Items | Provide Scrum Master Feedback on the story | |
| Provide Scrum Master Feedback on the story | Langchain Agent | Generate Scrum Master perspective feedback | Merge node | Update Scrum feedback in Backlog | |
| Update Scrum feedback in Backlog | Google Sheets | Write Scrum feedback back to backlog | Scrum Master feedback node | Loop Over Items for DoR check (cycle continuation) | |
| Loop Over Items Business Feedback | SplitInBatches | Batch processing for business feedback | Add Scrum Feedback to Story | OpenAI Business Analyst, Add Business feedback to Story | |
| OpenAI Business Analyst | LM Chat OpenAI | Generate business feedback AI responses | Loop Over Items Business Feedback | Business Validation | |
| Business Validation | Langchain Agent | Process business AI feedback | OpenAI Business Analyst | Update Business Validation in Backlog | |
| Update Business Validation in Backlog | Google Sheets | Write business feedback back to backlog | Business Validation | Loop Over Items Business Feedback (cycle continuation)| |
| Loop Over Items Technical Feedback | SplitInBatches | Batch processing for technical feedback | Add Business feedback to Story | OpenAI Technical Analyst, Add Technical feedback to the story | |
| OpenAI Technical Analyst | LM Chat OpenAI | Generate technical feedback AI responses | Loop Over Items Technical Feedback | Development Team Validation | |
| Development Team Validation | Langchain Agent | Process technical AI feedback | OpenAI Technical Analyst | Update Technical Validation in Backlog | |
| Update Technical Validation in Backlog | Google Sheets | Write technical feedback back to backlog | Development Team Validation | Loop Over Items Technical Feedback (cycle continuation)| |
| Add Scrum Feedback to Story | Merge | Combine Scrum feedback with story | Loop Over Items for DoR check | Loop Over Items Business Feedback, Add Business feedback to Story | |
| Add Business feedback to Story | Merge | Combine business feedback with story | Loop Over Items Business Feedback | Loop Over Items Technical Feedback, Add Technical feedback to the story | |
| Add Technical feedback to the story | Merge | Combine technical feedback with story | Loop Over Items Technical Feedback | Combine stories into one list | |
| Combine stories into one list | Aggregate | Aggregate all enriched stories into a single list | Add Technical feedback to the story | Draft Email for Refinement | |
| Draft Email for Refinement | Langchain Agent | Generate structured email content for refinement | Combine stories into one list | Ask Approval to Scrum Master | |
| OpenAI Scrum Master Emailer | LM Chat OpenAI | Assist in drafting email content | Combine stories into one list | Draft Email for Refinement | |
| Ask Approval to Scrum Master | Gmail | Send email requesting Scrum Master approval | Draft Email for Refinement | If Email content is Approved | |
| If Email content is Approved | If | Branch based on Scrum Master approval response | Ask Approval to Scrum Master | Send Email to Attendees, Create Draft in Scrum Master Email | |
| Send Email to Attendees | Gmail | Send finalized email to refinement attendees | If Email content is Approved | None | |
| Create Draft in Scrum Master Email| Gmail | Create draft email for manual adjustment | If Email content is Approved | Make aware the draft is ready to be adjusted manually | |
| Make aware the draft is ready to be adjusted manually | Gmail | Notify Scrum Master a draft is ready for review | Create Draft in Scrum Master Email | None | |
| Error Trigger | Error Trigger | Capture workflow errors | None | Send Error Email | |
| Send Error Email | Gmail | Send error notification email | Error Trigger | None | |
---
### 4. Reproducing the Workflow from Scratch
1. **Create Trigger Nodes:**
- Add a **Manual Trigger** node named “When clicking ‘Test workflow’” for manual runs.
- Optionally add a **Schedule Trigger** node (disabled by default) for periodic runs.
2. **Set Environment Variables:**
- Add a **Set** node named “Start Here: Set Environment Variables”. Define variables such as sheet names, backlog and DoR document IDs/links, statuses (e.g., "Ready for Refinement", "High Priority"), and any email templates or AI prompt parameters.
3. **Add No Operation Node:**
- Insert a **NoOp** node “No Operation, do nothing” connected from environment variables to allow parallel flow.
4. **Calendar Event Check:**
- Add a **Google Calendar** node “Load Calendar of Scrum Master” configured to use OAuth2 credentials to fetch events in a relevant time window, filtering on naming patterns related to refinement meetings.
- Connect **NoOp** node output to this calendar node.
5. **Check for Refinement Event:**
- Add a **Filter** node “Check for Refinement Project A” to detect if any calendar event matches the refinement criteria (e.g., event title).
- If no event, connect to a **NoOp** node “If no event do nothing” to end workflow path.
6. **Retrieve DoR and Backlog Links:**
- From “If no event do nothing” node, add two sequential **Google Drive** nodes:
- “Take DoR link for further referral” to get the DoR document link.
- “Take backlog link for further referral” to get backlog document link.
7. **Backlog Retrieval:**
- Add two **Google Sheets** nodes:
- “Select Stories Ready for Refinement from Backlog” to read stories with status “Ready for Refinement”.
- “Select Stories High Prio To Do from Backlog” to read stories marked as high priority and to-do.
- Both use the backlog link from the previous step.
8. **Merge Stories:**
- Add a **Merge** node “Join Potential Stories for Refinement” to combine these two story sets.
9. **Loop and Validate DoR:**
- Add a **SplitInBatches** node “Loop Over Items for DoR check” processing merged stories in batches.
- Connect one output to a **Google Sheets** node “Read DoR criteria” to fetch DoR criteria.
- Pass these to an **LM Chat OpenAI** node “OpenAI Scrum Master DoR Check” configured with a prompt to evaluate stories against each DoR criterium.
- Connect output to a **Langchain Agent** node “Compare User Story to DoR Criterium” for aggregation.
- Then use an **Aggregate** node “Aggregate DoR check to User Story Level” to consolidate results.
- Merge results back to the story data with a **Merge** node “Add DoR feedback to User Story”.
10. **Scrum Master Feedback:**
- Use a **Langchain Agent** node “Provide Scrum Master Feedback on the story” to generate Scrum Master insights on stories with DoR feedback.
- Update backlog sheet with a **Google Sheets** node “Update Scrum feedback in Backlog”.
- Loop back output to “Loop Over Items for DoR check” for continuous processing.
11. **Multi-Perspective AI Feedback:**
- Add a **SplitInBatches** node “Loop Over Items Business Feedback” connected from the Scrum feedback merge node.
- Connect it to an **LM Chat OpenAI** node “OpenAI Business Analyst” with business feedback prompt.
- Process output with a **Langchain Agent** “Business Validation”.
- Update backlog with “Update Business Validation in Backlog” Google Sheets node.
- Loop back to business feedback split node.
- Similarly, add a **SplitInBatches** node “Loop Over Items Technical Feedback” connected from business feedback merge node.
- Use “OpenAI Technical Analyst” LM Chat OpenAI node for technical feedback.
- Process with “Development Team Validation” Langchain Agent.
- Update backlog with “Update Technical Validation in Backlog” Google Sheets node.
- Loop back to technical feedback split node.
12. **Merge All Feedback:**
- Use a chain of **Merge** nodes to sequentially add Scrum, Business, and Technical feedback to stories:
- “Add Scrum Feedback to Story”
- “Add Business feedback to Story”
- “Add Technical feedback to the story”
- Aggregate with “Combine stories into one list”.
13. **Email Drafting and Approval:**
- Add a **Langchain Agent** node “Draft Email for Refinement” to create an HTML email body from combined stories.
- Assist with “OpenAI Scrum Master Emailer” LM Chat OpenAI node for tone and formatting.
- Use a **Gmail** node “Ask Approval to Scrum Master” to send email for approval with webhook enabled.
- Add an **If** node “If Email content is Approved” to branch on Scrum Master’s response.
- If approved, use **Gmail** node “Send Email to Attendees” to send the final email.
- If not approved, create a draft with **Gmail** node “Create Draft in Scrum Master Email”.
- Notify Scrum Master with “Make aware the draft is ready to be adjusted manually” Gmail node.
14. **Error Handling:**
- Add an **Error Trigger** node to catch workflow errors.
- Connect it to a **Gmail** node “Send Error Email” configured to notify admins with error details.
15. **Credentials Setup:**
- Configure Google Calendar, Google Drive, Google Sheets, and Gmail nodes with OAuth2 credentials.
- Configure OpenAI nodes with proper API key credentials.
16. **Testing and Validation:**
- Use manual trigger to test workflow end-to-end.
- Verify AI prompt outputs, email formatting, and error handling.
---
### 5. General Notes & Resources
| Note Content | Context or Link |
|-----------------------------------------------------------------------------------------------------------|--------------------------------------------------------------------------------------------------------|
| Use consistent event naming conventions in Google Calendar to ensure accurate refinement event detection. | Workflow setup instruction |
| Customize AI prompts in OpenAI nodes to align feedback tone and style with your team culture. | Workflow customization tip |
| Replace Google Sheets with Jira or Airtable by modifying data retrieval and update nodes accordingly. | Extensibility advice |
| Switch Gmail nodes to Outlook or SMTP nodes for email sending if preferred. | Integration flexibility |
| The workflow features retry and error handling with exponential backoff to improve robustness. | Operational resilience |
| For approval email webhook configuration, ensure the Gmail API is correctly set up with push notifications.| Gmail webhook integration details |
| Project credits and detailed blog post available at: https://example.com/automate-agile-refinement-prep | External resource for deeper understanding |
| AI-powered multi-perspective feedback is a unique selling point of this workflow, enhancing backlog quality.| Marketing point |
---
This completes the comprehensive reference for the "Automate Agile Refinement Prep with Gmail, OpenAI & Google Sheets" workflow.