AI Security & Analysis Toolkit

A set of lightweight, standard-library Python scripts designed to prepare data (codebases and network traffic) for AI analysis and security auditing.

Prerequisites: Python 3.x (No external libraries required).

1. Code Scraper (scraper.py)

Consolidates an entire Git repository or local directory into a single .txt file. Ideal for feeding context to LLMs.

Features:

• Automatically filters binaries and large files.
• Interactive prompt to exclude specific files by size.
• Supports GitHub URLs or local paths.

Usage:

python3 scraper.py
# Follow the interactive prompts to select source and exclusions.

---

2. HAR Cleaner (harcleaner.py)

Optimizes HAR (network log) files for AI analysis by removing "bloat" (images, fonts, binary blobs) to save tokens while preserving request logic.

Usage:

• Default (Best for Logic Analysis): Removes static assets, binary data, and timing metadata.

  python3 harcleaner.py traffic.har

• Keep CSS (For UI Analysis):

  python3 harcleaner.py traffic.har --keep-css

• Keep Everything (Just remove metadata):

  python3 harcleaner.py traffic.har --keep-static --keep-css --keep-binary

---

3. HAR Redactor (har_redact.py)

Surgically removes or replaces specific secrets (tokens, passwords, API keys) from HAR files without breaking the JSON structure.

Usage:

• Replace Secret: Finds the token and replaces it with [REDACTED].

  python3 har_redact.py traffic.har "MY_SECRET_TOKEN" --replace

• Delete Line: Removes the specific JSON key/value pair containing the secret.

  python3 har_redact.py traffic.har "MY_SECRET_TOKEN" --delete-line

• Delete Request: Removes the entire HTTP request/response entry if the secret is found anywhere inside it.

  python3 har_redact.py traffic.har "MY_SECRET_TOKEN" --delete-req

---

Quick Setup (Optional)

Make scripts executable to run them directly from your terminal:

chmod +x scraper.py harcleaner.py har_redact.py

Now you can run them as ./scraper.py, etc.