skills.rest
deonneon
deonneonCOMMUNITY
Explore all skills by this author →
GitHub

prompt-consultant

Parse Claude Code JSONL transcripts to extract and critique user prompts.

3|Updated Jan 25, 2026
Search Tags:#history#best-practices#claude-code#critique#prompting#prompt-analysis#prompt-audit
One-click install
npx skills add https://github.com/deonneon/agentic_automation_skills --skill prompt-consultant
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: prompt-consultant
Source: https://github.com/deonneon/agentic_automation_skills/tree/main/.claude/skills/prompt-consultant
Command: npx skills add https://github.com/deonneon/agentic_automation_skills --skill prompt-consultant

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Extracts all user prompts from Claude Code conversation history and delivers brutally honest critique to improve prompting practices.

Core Features & Use Cases

  • Copies Claude Code transcripts from ~/.claude/projects/
  • Parses JSONL files to extract user messages
  • Filters out tool results, system reminders, and other non-user content
  • Deduplicates and merges prompts for a concise critique
  • Delivers actionable feedback aligned with best practices

Quick Start

Feed Claude Code your prompt history and review the generated critique for immediate improvement.

Frequently Asked Questions about prompt-consultant

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I critique and improve prompts from Claude Code history?▼

To critique Claude Code prompts, this Skill parses JSONL transcripts from your project history, extracts user messages, filters out tool results, and deduplicates prompts to deliver actionable feedback on your prompting practices.

What is the best way to audit prompting practices across multiple Claude Code projects?▼

Auditing prompting practices across multiple projects involves merging JSONL conversation transcripts, filtering non-user content, and generating a rigorous critique to identify areas for structural prompt improvement.

Can I use my existing Claude Code history for a prompt audit without manual extraction?▼

Yes, you can use existing Claude Code history directly because the Skill copies JSONL transcript files from the local project directory, automatically isolating and deduplicating user messages for analysis.

Does the prompt critique process remove system reminders and tool results from transcripts?▼

Yes, the prompt critique process filters out system reminders, tool results, and other non-user content from JSONL transcripts to ensure the analysis focuses strictly on your actual user prompts.

How do teams get structured prompt improvements from Claude Code transcripts?▼

Teams get structured prompt improvements by feeding their Claude Code history into the analysis, which produces a rigorous critique with recommendations aligned with prompting best practices.

What are the limitations of analyzing Claude Code JSONL transcripts for prompt feedback?▼

Analysis is limited to user prompts extracted from local JSONL transcripts, meaning it cannot critique real-time conversational context or evaluate prompts not saved in the Claude Code project history files.

Related Skills

Explore More Content & Communication Skills→
kzarzyckikzarzycki

find-conversation

Search stored JSONL conversation history to locate past Claude Code sessions.

Community
Intermediate
cwijayasundaracwijayasundara
3

author-prompt-surface

Enforce Claude Code prompting standards for agent, skill, and command prompts.

Community
Intermediate
kudrmaxkudrmax
1

search-conversations

Search past Claude Code sessions across projects using keyword queries.

Community
Intermediate
kwoekelkwoekel

prompting-pattern-library

Analyze prompt quality and provide prompting patterns for LLM interactions.

Community
Advanced