Lennys-to-sop-and-skills

🥈Silver

Transforms Lenny's Podcast transcripts into professional SOPs and Claude Skills. Benefits operations teams by automating knowledge extraction and documentation. Connects to Claude for skill generation and integrates with internal knowledge bases.

1680Updated 1mo ago

Intermediate30min to implementautomation

Saves ~120 min per use

Quick InstallView Source

git clone https://github.com/qingxuantang/Lennys-to-sop-and-skills.git

Works with:

Claude

Overview

About This Skill

How to Use

[{"step":"Prepare the transcript. Copy the Lenny's Podcast transcript into a clean text file or paste it directly into your AI tool. Ensure the transcript is complete and includes timestamps if available.","tip":"For best results, use transcripts from episodes focused on operational topics (e.g., \"How to Build Scalable Systems\" or \"The Art of Documentation\"). Avoid episodes with heavy storytelling or minimal actionable content."},{"step":"Run the prompt. Use the provided prompt template in your AI tool (e.g., Claude, ChatGPT) and replace [PASTE_TRANSCRIPT_HERE] with your transcript. Specify if you want the output in Markdown, JSON, or another format.","tip":"If the transcript is long (>5,000 words), split it into sections (e.g., by topic) and process each section separately to avoid overwhelming the AI."},{"step":"Refine the output. Review the generated SOP and Claude Skill definition. Add or remove steps based on your team's workflow. For example, if your team uses Jira for tracking, include a step to create Jira tickets for action items.","tip":"Use the example output as a reference for structure. Pay attention to the [ACTIONS], [DECISION POINTS], and [OUTPUTS] labels to ensure clarity."},{"step":"Integrate with tools. If you're generating a Claude Skill, copy the definition into your Claude Skills library. For SOPs, upload the document to your internal knowledge base (e.g., Notion, Confluence) and share it with your team.","tip":"For Claude Skills, test the skill with a small dataset first. Use the `debug` mode to identify any issues with triggers or outputs."},{"step":"Iterate and improve. Gather feedback from your team on the SOP or skill. Update the documentation based on their input, and refine the AI-generated content to better fit your workflows.","tip":"Schedule a review session 2 weeks after implementation to assess adoption and identify gaps. Use this feedback to retrain the AI or adjust the prompt template."}]

Setup & Installation

Quick Install

No install command available. Check the GitHub repository for manual installation instructions.

Alternative Install (Git Clone)

git clone https://github.com/qingxuantang/Lennys-to-sop-and-skills

Requirements

Claude Code or compatible AI agent
Works with: Claude

Quick Start Guide

Install the Skill

Copy the install command above and run it in your terminal.

Open Your AI Agent

Launch Claude Code, Cursor, or your preferred AI coding agent.

Try It Out

Use the prompt template or examples below to test the skill.

Customize

Adapt the skill to your specific use case and workflow.

Usage Examples

Prompt Template

Convert the following Lenny's Podcast transcript into a detailed Standard Operating Procedure (SOP) and a Claude Skill definition. Focus on extracting actionable insights, step-by-step workflows, and key takeaways. Include: 
1. A clear title and purpose for the SOP/Skill.
2. Step-by-step instructions with [ACTIONS], [DECISION POINTS], and [OUTPUTS].
3. Key metrics or KPIs to track success.
4. A Claude Skill definition with triggers, inputs, and outputs.

Transcript: [PASTE_TRANSCRIPT_HERE]

Example Output

### SOP: Automating Customer Feedback Analysis with NLP
**Purpose:** Standardize the process of analyzing customer feedback using NLP tools to identify trends, sentiment, and actionable insights for product improvements.

**Prerequisites:** 
- Access to customer feedback dataset (e.g., Zendesk, Intercom, or survey responses)
- NLP tool (e.g., MonkeyLearn, Lexalytics, or custom Python script)
- Output template for reporting (Google Sheets or Notion)

**Steps:**
1. **Data Collection [ACTION]:** Aggregate feedback from all sources into a single CSV/JSON file. Clean data by removing duplicates and irrelevant entries (e.g., spam).
   - *Decision Point:* If feedback volume exceeds 10,000 entries, use a sampling method (e.g., 10% random sample) to reduce processing time.
   - *Output:* `cleaned_feedback_dataset.csv`

2. **Sentiment Analysis [ACTION]:** Run the dataset through an NLP tool to classify sentiment (positive, neutral, negative) and extract key themes (e.g., "pricing," "usability," "support").
   - *Tool Example:* Use MonkeyLearn’s pre-trained sentiment analyzer with a confidence threshold of 0.7.
   - *Output:* `sentiment_analysis_results.json` with sentiment scores and theme frequencies.

3. **Trend Identification [ACTION]:** Group feedback by theme and time period (e.g., weekly/monthly) to identify trends. Calculate the percentage of negative feedback per theme.
   - *Decision Point:* Flag themes with >30% negative sentiment for immediate review.
   - *Output:* `trend_report.csv` with theme, sentiment distribution, and trend direction (improving/worsening/stable).

4. **Actionable Insights [ACTION]:** For each flagged theme, extract specific quotes or examples to support findings. Assign ownership to teams (e.g., Product, Support) for follow-up.
   - *Output:* `insights_report.md` with:
     - Top 3 themes by negative sentiment
     - Example quotes
     - Recommended actions (e.g., "Investigate checkout flow for usability issues")

5. **Reporting [ACTION]:** Compile results into a dashboard (e.g., Google Data Studio) and share with stakeholders weekly. Include:
   - Sentiment score over time
   - Top themes by volume and sentiment
   - Action items with owners and deadlines

**KPIs:**
- Average sentiment score (target: >2.5/5)
- Time to resolution for flagged themes (<7 days)
- Percentage of feedback processed weekly (>80%)

### Claude Skill Definition
**Name:** Customer Feedback Analyzer
**Description:** Automates the analysis of customer feedback to identify sentiment trends, themes, and actionable insights.

**Triggers:**
- New feedback dataset uploaded to [INPUT_FOLDER]
- Scheduled weekly analysis (every Monday at 9 AM)

**Inputs:**
- `feedback_dataset.csv` (raw feedback data)
- `nlp_tool_config.json` (API keys, model settings)

**Outputs:**
- `sentiment_analysis_results.json`
- `trend_report.csv`
- `insights_report.md`
- Slack notification to #feedback-alerts channel with summary

**Error Handling:**
- If sentiment analysis fails, retry with a smaller dataset or notify admin.
- Log all errors to `feedback_analysis_errors.log`.