Back to Prompts
Content GenerationClaude Sonnet 4.5

Generate Standard Operating Procedure

Part of the SOP Generator template

Prompt
Create a standard operating procedure (SOP) document for the following process.

Process Name:
{{process_name}}

Process Description:
{{process_description}}

Target Audience:
{{audience}}

Create the SOP in this format:

# {{process_name}}

**Version:** 1.0
**Last Updated:** [Today's date]
**Owner:** [Role responsible for this SOP]
**Applies To:** {{audience}}

## Purpose
[1-2 sentences explaining why this process exists and what it accomplishes]

## Scope
- **In scope:** [What this SOP covers]
- **Out of scope:** [What this SOP does not cover]

## Prerequisites
- [What must be in place before starting]
- [Required access, tools, or permissions]

## Definitions
| Term | Definition |
|------|------------|
| [Term] | [Definition] |

## Procedure

### Step 1: [Action Title]
**Who:** [Role]
**When:** [Trigger or timing]

1. [Specific action]
2. [Specific action]
3. [Specific action]

**Expected outcome:** [What should be true when this step is complete]

[Continue for all steps...]

## Decision Points
- **If [condition]:** [what to do]
- **If [other condition]:** [what to do]

## Exceptions & Escalation
- [When to deviate from standard process]
- [Who to escalate to and when]

## Quality Checks
- [ ] [Checkpoint 1]
- [ ] [Checkpoint 2]

## Related Documents
- [Link to related SOPs or resources]

## Revision History
| Version | Date | Author | Changes |
|---------|------|--------|---------|
| 1.0 | [Date] | [Author] | Initial version |

---

Guidelines:
- Each step should be a single, clear action
- Use active voice ("Click the button" not "The button should be clicked")
- Include screenshots or examples where helpful (note where they should go)
- Specify roles, not names (people change, roles persist)
- Include approximate time for time-sensitive steps
Variables
Replace these placeholders with your specific values
{{process_name}}

Name of the process

Example:

New Client Onboarding

{{process_description}}

Description of what the process involves

Example:

Setting up a new client in our systems, including CRM entry, n8n instance provisioning, and kickoff call scheduling

{{audience}}

Who will follow this SOP

Example:

Account Managers and Operations team

Example

Input

Process: Handling a failed n8n workflow alert. Description: When we get an alert that a client workflow has failed, we need to investigate, fix or escalate, and notify the client. Audience: Support team members

Output

# Handling Failed Workflow Alerts

**Version:** 1.0
**Last Updated:** January 2024
**Owner:** Support Team Lead
**Applies To:** Support team members

## Purpose
Ensure failed client workflows are quickly investigated and resolved with appropriate client communication, minimizing downtime and maintaining client trust.

## Scope
- **In scope:** All automated workflow failure alerts from the monitoring system
- **Out of scope:** Planned maintenance windows, client-requested pauses

## Prerequisites
- Access to monitoring dashboard
- Access to client n8n instances
- Ability to send client communications

## Definitions
| Term | Definition |
|------|------------|
| Critical workflow | Revenue-impacting or client-facing automation |
| Silent failure | Workflow fails without error notification to client |

## Procedure

### Step 1: Acknowledge the Alert
**Who:** On-call support
**When:** Within 15 minutes of alert

1. Open the alert in the monitoring dashboard
2. Click "Acknowledge" to prevent re-alerting
3. Note the client name, workflow name, and error type

**Expected outcome:** Alert marked as acknowledged, initial context gathered

### Step 2: Assess Severity
**Who:** On-call support
**When:** Immediately after acknowledgment

1. Check if this is a critical workflow (see client tier in CRM)
2. Determine if data was lost or if this is a temporary failure
3. Check if workflow has auto-retry enabled

**Expected outcome:** Severity level determined (Critical/High/Medium/Low)

### Step 3: Investigate Root Cause
**Who:** On-call support
**When:** Within 30 minutes for Critical/High, 2 hours for Medium/Low

1. Access the client's n8n instance
2. Open the failed execution
3. Review the error message and stack trace
4. Check if this is a recurring issue (look at recent executions)
5. Document findings in the ticket

**Expected outcome:** Root cause identified or escalation path determined

[Additional steps would continue...]

## Decision Points
- **If the fix is straightforward:** Implement fix, test, and close
- **If root cause is unclear:** Escalate to engineering within 1 hour
- **If client action required:** Notify client before investigating further

## Quality Checks
- [ ] Client notified if downtime > 1 hour
- [ ] Root cause documented in ticket
- [ ] Fix verified with successful test execution
- [ ] Monitoring confirmed for next 24 hours