Uday Lunawat udaylunawat

AI Harness Feature Inventory and Failure Modes

Purpose: A practical feature taxonomy for evaluating or designing an AI harness around models such as ChatGPT, Codex, Claude Code, Cursor, OpenCode, Gemini CLI, Cline/Roo Code, or a custom agent runtime.

The key idea is that a harness should not be evaluated only by asking "What can the model do?" It should also be evaluated by asking:

What happens to the user when a harness feature is missing?

A capable model can still feel unusable if the harness has poor state management, no recovery paths, weak context controls, insufficient execution isolation, or no observability.

GitHub Gist Files:

Script File: telegram_rental_scraper.py

Contains the main script provided.

Requirements File: requirements.txt

telethon
pandas

(async function executeInstagramPostCollection(maxPosts = 10) {
    console.log(`Starting Instagram post collection (limit: ${maxPosts} posts)...`);

    // Verify we're on Instagram
    if (!window.location.hostname.includes('instagram.com')) {
        console.error("This script should be run on Instagram.com");
        return;
    }

Breathe - text animation

A simple breathing text animation.

A Pen by Tanner Dolby on CodePen.

License.

	#!/usr/bin/env python3
	import argparse
	import json
	import csv
	import os
	import sys
	import time
	import requests
	from bs4 import BeautifulSoup
	import concurrent.futures

	#!/usr/bin/env python3
	import argparse
	import json
	import csv
	import sys
	csv_max = sys.maxsize
	while True:
	try:
	csv.field_size_limit(csv_max)
	break

	import os
	os.environ["TF_CPP_MIN_LOG_LEVEL"] = "3"
	import gc
	import subprocess
	import random
	import numpy as np
	import pandas as pd
	import seaborn as sns
	import matplotlib.pyplot as plt

	import os
	os.environ["TF_CPP_MIN_LOG_LEVEL"] = "3"
	import gc
	import subprocess
	import random
	import numpy as np
	import pandas as pd
	import seaborn as sns
	import matplotlib.pyplot as plt

	# https://stackoverflow.com/a/64006242/9292995
	# https://github.com/jfilter/split-folders
	import splitfolders

	# If your datasets is balanced (each class has the same number of samples), use ratio
	# otherwise use fixed if dataset is imbalanced.
	splitfolders.ratio('input_dir', output="output_dir", oversample=False, ratio = (0.8, 0.1, 0.1),
	seed=1337)

	# https://www.tensorflow.org/datasets/api_docs/python/tfds/folder_dataset/ImageFolder