From memorylane
Generates step-by-step automation runbooks (.md files) from workflow patterns by analyzing activity timelines with search and browse tools. Useful for scripting manual processes after /discover-patterns.
npx claudepluginhub deusxmachina-dev/memorylane --plugin memorylaneThis skill is limited to using the following tools:
Generate a step-by-step automation runbook (`.md` file) from a workflow pattern. The pattern can come from a previous `/discover-patterns` session or be described directly by the user.
Generates downloadable PDF process briefing from detected patterns or workflows: visual flowchart, steps with apps/actions, stats, and automation suggestions. Useful for documenting real team processes.
Provides proven architectural patterns for n8n workflows: webhook processing, HTTP API integration, database operations, AI agents, batch processing, scheduled tasks. Use when building, designing, or automating with n8n.
Provides guidance on AI tools for automating repetitive tasks, Zapier integrations, no-code workflows, and RPA. Activates when automation, Zapier, or workflow keywords are mentioned.
Share bugs, ideas, or general feedback.
Generate a step-by-step automation runbook (.md file) from a workflow pattern. The pattern can come from a previous /discover-patterns session or be described directly by the user.
/discover-patterns run: use the pattern name and details they shared. Proceed to Step 2./discover-patterns first to find candidates.Use the timeline and activity tools to build a complete picture of the workflow:
search_context(query) — search for activities matching the pattern across 30 days to find all instances.browse_timeline — for the strongest matches, scan the surrounding timeline to capture the full workflow sequence (steps before and after the core actions).get_activity_details(ids) — for at least 3 clear instances, extract exact steps, apps, URLs, UI elements, and data flow.The goal is to reconstruct the complete process from trigger to completion, and to understand what varies vs. what stays the same across runs.
From the evidence, map the complete workflow from trigger to completion:
Cross-reference multiple instances to reconstruct the complete sequence. Then separate:
For each step, also consider error points:
Using the evidence gathered, produce the runbook following the Output Template below. Ensure:
Ask the user where to save the runbook file. Suggest a default:
~/Desktop/runbooks/[pattern-name-slug].md
Where [pattern-name-slug] is the pattern name lowercased with spaces replaced by hyphens (e.g., "Client Onboarding" → client-onboarding.md).
Save the runbook to the user's chosen path. Create the directory if it doesn't exist.
After saving, show:
/pattern-to-runbook again for another pattern# [Pattern Name] — Automation Runbook
## Overview
- **What this does**: [1-2 sentence description of the end-to-end process]
- **Trigger**: [what starts it — time-based, event-based, or manual]
- **Frequency**: [how often it occurs, based on activity evidence]
- **Estimated time per run**: [based on activity timestamps]
- **Estimated time saved per week**: [frequency × time per run]
## Prerequisites
- **Apps/services**: [list each app or service with what it's used for]
- **Access needed**: [credentials, API keys, permissions — don't include actual secrets]
- **Input data**: [what data sources feed into this process]
## Steps
### 1. [Action verb] — [what happens] in [app]
- **Details**: [exactly what to do]
- **Input**: [what data goes in]
- **Output**: [what to expect / what gets produced]
- **Error handling**: [what can go wrong and what to do]
### 2. [Action verb] — [what happens] in [app]
- **Details**: ...
- **Input**: ...
- **Output**: ...
- **Error handling**: ...
[Continue for all steps]
## What Varies Between Runs
- [Parameter 1]: [description and example values from evidence]
- [Parameter 2]: ...
These are the inputs/arguments for any automation built from this runbook.
## What Stays Constant
- [Constant 1]: [value or description]
- [Constant 2]: ...
These get hardcoded in the automation.
## Error Handling
- **[Failure mode 1]**: [how to detect] → [what to do]
- **[Failure mode 2]**: [how to detect] → [what to do]
## Automation Approach
- **Recommended method**: [API script / browser automation / CLI tool / scheduled job / etc.]
- **Key APIs or services**: [specific APIs, webhooks, or integrations to use]
- **Implementation sketch**:
[Pseudocode or high-level steps for the automation]
- **Effort estimate**: [easy / medium / hard] — [brief justification]
Before finalizing the runbook, verify:
get_activity_details may contain sensitive information. Extract only the process-relevant details (app names, field labels, URLs). Never include passwords, API keys, or personal messages in the runbook.