Does this skill log sensitive information?

Yes, it logs all prompts and tool outputs locally in unencrypted markdown files. Users should be mindful of what they share and use the pruning scripts regularly.

Can I use these logs to give Claude context from yesterday?

Absolutely. You can use the @ symbol in Claude Code to reference a specific markdown log file, allowing Claude to read and understand previous work history.

How does it handle multiple simultaneous Claude sessions?

The skill uses unique session IDs to create separate log files for each instance, preventing data collisions and ensuring chronological accuracy.

Is any external software required for the logging to work?

The script works with standard bash, but installing 'jq' is highly recommended to ensure the JSON event data is parsed accurately.

Where are the logs stored on my system?

By default, logs are stored in the ~/.claude/conversation-logs directory, though this can be customized within the log-conversation.sh script.

Conversation Logging

Name: Conversation Logging
Author: ianphil

byianphil

Analytics & Monitoring

Automatically logs Claude Code interactions, prompts, and tool executions to organized markdown files for auditing and session recovery.

About

The Conversation Logging skill implements global hooks to capture every interaction within Claude Code, creating a persistent record of prompts, responses, and tool outputs. By organizing logs by session ID into structured markdown files, it enables developers to debug failed commands, audit AI-driven changes, and seamlessly provide historical context to Claude when resuming work. This skill is essential for users who need a searchable history of their AI-assisted development workflow or a reliable way to pick up exactly where they left off in previous sessions.

Key Features

Detailed tool execution tracking for Bash, Read/Write, and WebSearch
Structured markdown logging of all prompts and responses
0 GitHub stars
Built-in log pruning scripts to manage local storage and delete old entries
Easy context recovery by referencing previous log files using the @ symbol
Session-based organization to prevent logs from overlapping in concurrent instances

Use Cases

Debugging failed tool executions or complex bash command sequences
Auditing AI-generated code changes for security and compliance
Resuming long-running development tasks across different days or terminal sessions

About

Key Features

Detailed tool execution tracking for Bash, Read/Write, and WebSearch
Structured markdown logging of all prompts and responses
0 GitHub stars
Built-in log pruning scripts to manage local storage and delete old entries
Easy context recovery by referencing previous log files using the @ symbol
Session-based organization to prevent logs from overlapping in concurrent instances

Use Cases

Debugging failed tool executions or complex bash command sequences
Auditing AI-generated code changes for security and compliance
Resuming long-running development tasks across different days or terminal sessions