What metadata is included in the agent report?

The report includes the agent slug, the unique agent ID, the specific AI model used, a timestamp, and the complete text of the final response.

Where does the skill save the generated reports?

By default, reports are saved as Markdown files in your current working directory using the naming convention 'agent-[ID]-report.md'.

Can I use this to see what a subagent produced?

Yes, this skill is specifically designed to isolate and display the final message from subagents involved in complex workflows.

Do I need to provide a file path for the logs?

No, the skill uses an internal script to locate and parse the relevant JSONL logs based on the agent ID you provide.

What is a Claude agent ID?

A Claude agent ID is a unique short identifier, such as 'ad42ecb', assigned to a specific subagent or task execution within the Claude environment.

Agent Report Extractor

Name: Agent Report Extractor
Author: OhadRubin

byOhadRubin

•

Analytics & Monitoring

Extracts and formats final messages from Claude agent execution logs into clean, readable Markdown reports.

The Agent Report Extractor is designed to improve visibility into multi-agent workflows by automatically parsing JSONL log files. Instead of manually searching through raw data, this skill allows users to quickly retrieve the final response, model metadata, and timestamps from a specific agent ID. It is an essential tool for developers and AI researchers who need to audit agent performance, debug subagent interactions, or document the specific outputs of an automated coding session.

Key Features

01Supports quick-access via short agent IDs

02Captures critical metadata including model name and timestamps

03Generates structured Markdown reports for easy readability

041 GitHub stars

05Automated extraction of final messages from JSONL logs

06Optional custom output paths for report generation

Use Cases

01Reviewing the final output of a complex multi-agent coding task

02Generating clean documentation of agent-driven results for project records

03Auditing subagent responses to debug logic or chain-of-thought issues

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add ohadrubin/skills agent-report

For use in Claude.ai and ChatGPT

Download Skill