# n8n automation project brief

Use invented sample data. Keep passwords, API tokens and unnecessary personal data out of this document.

## 1. Business problem
- Task performed today:
- Who performs it:
- Frequency and volume:
- Time spent or common errors:

## 2. First useful outcome
- What should happen:
- What is outside the first version:
- How success will be measured:

## 3. Trigger and connected systems
- Source system and trigger:
- Destination system and supported API:
- Owner of each system:
- Test environment and production environment:
- Authentication approach (do not include secrets):

## 4. Data contract
- Required input fields:
- Synthetic input example:
- Destination field mapping:
- Stable source identifier / idempotency key:
- Sensitive fields and retention period:

## 5. Rules and optional AI
- Deterministic business rules:
- Is AI necessary? Why:
- Allowed AI output and validation:
- How uncertain or invalid output is handled:
- Actions requiring human approval:

## 6. Reliability and recovery
- Durable duplicate-prevention approach:
- Timeout-after-acceptance behaviour:
- Retry limits and backoff:
- Failure owner and escalation route:
- How missing events or missed schedules are detected:
- Reversal or recovery procedure:

## 7. Acceptance checks
- Valid request:
- Missing input:
- Duplicate / concurrent duplicate:
- Timeout after a destination accepted the request:
- Destination unavailable:
- Unexpected AI output:
- Approval declined:
- Expired credential:

## 8. Delivery and ownership
- Desired timeline:
- Hosting, service and optional AI costs:
- Documentation and handover:
- Ongoing monitoring and maintenance owner:
- Who approves release:

This is a planning document, not an executable n8n workflow.
Guide: https://mfaruk.com/blog/n8n-automation-project-brief
Project enquiries: https://mfaruk.com/contact