What prerequisites are needed to use the save-transcript skill?

You must have a Notion database set up with specific properties (Title, Tool, Model, Tags) and an active Notion MCP server configured with access to that database.

How do I configure my Notion database for this skill?

You can provide your Notion database URL directly to the assistant during the first run, and it will extract the ID and update the configuration automatically.

Can this skill be used with Cursor or other CLI tools?

Yes, it is compatible with Claude Code, Cursor, and Codex CLI, and it will automatically attempt to detect and log which tool is currently in use.

Does it save long tool outputs like file contents or terminal logs?

The skill intelligently summarizes tool results—such as file reads or bash outputs—to keep the transcript concise while preserving the verbatim text of the user and assistant messages.

Notion Conversation Exporter

Name: Notion Conversation Exporter
Author: wustep

bywustep

Productivity & Workflow

Exports Claude Code and Cursor conversation transcripts directly to a structured Notion database for documentation and archival.

About

The save-transcript skill automates the process of capturing and archiving AI interactions by transforming chat histories into organized Notion pages. It intelligently parses conversation context to generate concise titles, detect model metadata, and categorize sessions with relevant tags. By formatting messages into Notion-flavored markdown—including summaries, tool call logs, and verbatim message preservation—it creates a searchable, professional-grade knowledge base of your AI-assisted development workflows without manual effort.

Key Features

Dynamic tag assignment based on conversation content analysis
Verbatim transcript preservation using Notion callout blocks for readability
Intelligent metadata detection for AI models and development tools
0 GitHub stars
Automatic session summarization highlighting key decisions and outcomes
Condensed tool-call logging to capture execution context without bloat

Use Cases

Documenting AI-assisted code refactoring for team reviews and audit trails
Creating a searchable knowledge base of architectural decisions and debugging sessions
Archiving complex prompt-engineering workflows for future reference

About

Key Features

Dynamic tag assignment based on conversation content analysis
Verbatim transcript preservation using Notion callout blocks for readability
Intelligent metadata detection for AI models and development tools
0 GitHub stars
Automatic session summarization highlighting key decisions and outcomes
Condensed tool-call logging to capture execution context without bloat

Use Cases

Documenting AI-assisted code refactoring for team reviews and audit trails
Creating a searchable knowledge base of architectural decisions and debugging sessions
Archiving complex prompt-engineering workflows for future reference