Where are the investigation logs stored?

Logs are stored as markdown files within the performance subdirectory of your AI_STATE_DIR, which defaults to .claude.

Why are verbatim user quotes required?

Verbatim quotes ensure the investigation remains focused on the exact problem reported by the user without misinterpretation during the logging process.

What is the primary purpose of this skill?

It provides a structured way to document performance investigations, ensuring every optimization is supported by evidence and clear rationale.

How does this help with long-term maintenance?

By creating standardized logs of profiling results and decisions, teams can easily review the evidence behind past changes and track performance regressions over time.

Does it require a specific document to function?

Yes, it follows docs/perf-requirements.md as the canonical contract for how investigations should be recorded.

Performance Investigation Logger

Name: Performance Investigation Logger
Author: agent-sh

byagent-sh

•

분석 및 모니터링

Logs structured performance investigation notes and evidence to maintain a rigorous, auditable optimization workflow.

The Performance Investigation Logger is designed to standardize the way performance-related inquiries are documented within the AI workflow. It ensures that every phase of an investigation—from initial user quotes to final decisions—is backed by hard evidence such as specific command outputs, file paths, and metrics. By enforcing a consistent markdown structure and utilizing the AI state directory, it creates a searchable, persistent record of performance bottlenecks and the rationale behind optimization choices, facilitating better debugging and long-term project maintenance.

주요 기능

01Strict enforcement of canonical performance requirements and contracts

021 GitHub stars

03Mandatory evidence linking for metrics, commands, and files

04Phase-based documentation of performance optimization steps

05Structured investigation logging with verbatim user quotes

06Integration with AI_STATE_DIR for persistent state management

사용 사례

01Recording baseline benchmarks and subsequent evidence for speed improvements

02Creating an audit trail for performance-related refactors and optimizations

03Tracking and documenting the root cause analysis of application latency

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add agent-sh/perf perf-investigation-logger

For use in Claude.ai and ChatGPT

Download Skill