Does it require specific file formats?

No, it can analyze both raw text and markdown content by looking for internal patterns and structural markers.

What happens if the classification confidence is low?

If the confidence score falls below 70%, the skill defaults to '_untyped' and flags the log for manual review to prevent incorrect categorization.

Can it reclassify files that are already logged?

Yes, the skill includes a reclassification operation that allows you to update the type of existing files, even if they were previously categorized.

How does it determine the log type?

It uses a scoring system (0-100) based on keyword frequency, command context (like 'npm build' or 'pytest'), file structure, and metadata like exit codes.

What log types can this skill identify?

It identifies 10 specific types: session, build, deployment, debug, test, audit, operational, changelog, workflow, and untyped.

Log Classifier

Name: Log Classifier
Author: fractary

byfractary

•

分析と監視

Categorizes log files into specific types using advanced pattern matching, keyword analysis, and structural metadata.

概要

The Log Classifier skill enables Claude to intelligently identify and organize various types of logs, including build, test, deployment, and audit records. By analyzing content signals such as exit codes, command contexts, and path patterns, it assigns logs to one of ten supported categories with a calculated confidence score. This skill is essential for developers and DevOps engineers who need to automate the processing of diagnostic data, troubleshoot pipeline failures, or maintain structured documentation across complex software environments.

主な機能

Analyzes frontmatter, metadata, and command-line context
Calculates confidence scores based on keyword and pattern matches
Provides detailed reasoning for every classification decision
Supports reclassification of existing untyped log files
2 GitHub stars
Identifies 10 distinct log types including Build, Test, and Audit

ユースケース

Automating the categorization of untyped logs in a CI/CD pipeline
Differentiating between application errors and environment deployment failures
Organizing project workspaces by automatically tagging session and test logs

概要

主な機能

Analyzes frontmatter, metadata, and command-line context
Calculates confidence scores based on keyword and pattern matches
Provides detailed reasoning for every classification decision
Supports reclassification of existing untyped log files
2 GitHub stars
Identifies 10 distinct log types including Build, Test, and Audit

ユースケース

Automating the categorization of untyped logs in a CI/CD pipeline
Differentiating between application errors and environment deployment failures
Organizing project workspaces by automatically tagging session and test logs