How do I find my current Claude Code session ID?

The skill automatically utilizes the CLAUDE_SESSION_ID environment variable to locate the correct log file within the workspace configuration directory.

How does this complement the learn-from-mistakes skill?

While get-history provides the raw data and logs of what occurred, learn-from-mistakes uses that data to perform high-level analysis and recommend behavioral fixes.

Does the get-history skill modify my conversation data?

No, this skill provides read-only access to your session storage for analysis and debugging purposes; it does not alter the history files.

Can I see logs from agent sidechains using this skill?

Yes, the skill is designed to identify and access sidechain logs using the agent-{agent-id}.jsonl naming pattern in your project workspace.

What data format is the conversation history stored in?

History is stored in JSONL (JSON Lines) format, which includes structured entries for summaries, user/assistant messages, tool invocations, and tool results.

Claude Session History

Name: Claude Session History
Author: cowwoc

bycowwoc

•

分析と監視

Accesses and parses raw Claude Code conversation history to analyze agent behavior and debug tool call sequences.

概要

The Get History skill provides developers with direct access to the raw conversation data stored during a Claude Code session. By retrieving JSONL logs from the workspace session storage, it allows for deep inspection of tool invocations, user-assistant exchanges, and agent sidechain processes. This capability is essential for debugging unexpected agent outputs, verifying the chronological sequence of tool results, and optimizing agent-driven workflows through precise analysis of past interactions via automated parsing tools like jq.

主な機能

Direct access to raw session JSONL logs via session ID
Support for both main session and agent sidechain logs
Seamless integration with error analysis and debugging workflows
2 GitHub stars
Automated parsing of tool calls and results using jq
Chronological extraction of messages and conversation summaries

ユースケース

Analyzing the decision-making process of autonomous agent sidechains
Verifying the exact content and metadata of past assistant responses
Debugging unexpected tool behavior or infinite loop patterns

概要

主な機能

Direct access to raw session JSONL logs via session ID
Support for both main session and agent sidechain logs
Seamless integration with error analysis and debugging workflows
2 GitHub stars
Automated parsing of tool calls and results using jq
Chronological extraction of messages and conversation summaries

ユースケース

Analyzing the decision-making process of autonomous agent sidechains
Verifying the exact content and metadata of past assistant responses
Debugging unexpected tool behavior or infinite loop patterns