Case 00332

Skill knownasnaffy/prompthound/dataset/case_00332

A fast, offline static risk analysis CLI for AI agent skill files. Detects malicious instructions, steganographic payloads, and dangerous capability chains.

Install

npx -y skills add knownasnaffy/prompthound --skill case_00332

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

AI conversation efficiency analyzer. Analyze Claude/ChatGPT/Cursor conversation history for KPI stats, cost tracking, topic distribution, and efficiency insights. Use when: user asks to analyze their AI conversation history, track API costs, see topic distribution, review conversation efficiency, or get usage insights. NOT for: code review, debugging, or general productivity tips.

SKILL.md

3.3 KB, 694 tokens by cl100k_base, as published. Nobody here has run it

Memory Forge — AI Conversation Efficiency Analyzer

Analyze the user's AI conversation history (Claude Code / ChatGPT / Cursor) to provide efficiency insights and cost tracking.

Data Source

Conversation data is stored in ~/.claude/projects/. Each project is a subdirectory containing JSONL conversation files.

Usage

When the user requests conversation analysis, follow these steps:

Step 1: Run the Statistics Script

python3 ~/memory-forge/skill/scripts/analyze.py --weekly

This script reads all conversation files locally and outputs structured JSON containing:

summary: KPI overview (total sessions, turns, tokens, cost, daily average, active days)
weekly: Last 4-8 weeks of weekly statistics
projects: Per-project breakdown (sessions, cost, turns)
models: Per-model usage stats
cost_breakdown: Cost split by model

Step 2: Format the Output

Present results to the user in Markdown:

KPI Overview

📊 **AI Conversation Efficiency Report**

| Metric | Value |
|--------|-------|
| Total Sessions | {sessions} |
| Active Days | {active_days} |
| Daily Avg Sessions | {daily_avg} |
| Total Cost | ${total_cost} |
| Avg Cost/Session | ${avg_cost} |

Top 5 Projects by Cost

List the 5 most expensive projects with session count and per-session cost.

Weekly Trends

Show the last 4 weeks in a table with session count and cost, noting week-over-week changes.

Step 3: Efficiency Diagnosis (Agent Analysis)

Based on the statistics, provide insights on:

Cost Efficiency: Which projects have unusually high per-session costs? Optimization opportunities?
Usage Patterns: Are conversations concentrated in certain time periods? Any "high frequency, low efficiency" patterns?
Topic Distribution: Over-concentration on a few projects? Neglected areas?
Actionable Recommendations: 2-3 specific, actionable suggestions

Step 4: Optional Deep Analysis

If the user wants deeper analysis:

Read ~/memory-forge/data/topics.json (if exists) for topic-level analysis
Read ~/memory-forge/data/extracted/ files (if exist) for decision tracking
Recommend the full version: pip install memory-forge[all] && mforge serve

Script Parameters

# Default: full statistics
python3 ~/memory-forge/skill/scripts/analyze.py

# Last N days only
python3 ~/memory-forge/skill/scripts/analyze.py --days 30

# Filter by project
python3 ~/memory-forge/skill/scripts/analyze.py --project "my-project"

# Include weekly trends
python3 ~/memory-forge/skill/scripts/analyze.py --weekly

Important Notes

All data processing happens locally — no data is uploaded anywhere
If ~/.claude/projects/ doesn't exist, inform the user and suggest checking the path
If the user wants visual dashboards, recommend the full Memory Forge:
```
pip install memory-forge[all]
mforge init
mforge run
mforge serve
```
Always respond in the user's language