pliablepixels

## 👤 User

Analyze agents.md to make sure its in great shape. One key issue I see is you implement things, run tests, say its all ok, but it breaks when I try. Are you generating/updating tests always? If not, how shoulD AGENTS.md be improved? I really want to follow a "test for everything" philosophy - when tests pass, it really should mean the app should work and all options/buttons/screens tested

---

## 🤖 Claude

I'll analyze AGENTS.md and identify the testing gaps. Let me first check what tests currently exist:

pliablepixels

## 👤 User

The full walkthrough tests are surfacial. If I want to write human readable instructions on the flow (I want it to go step by step into each screen and test functionality of items), and use that human redable instructions to generate code for e2e tests, what is the best approach?

---

## 🤖 Claude

Great question! For human-readable test specifications that generate E2E tests, here are the best approaches for your Playwright setup:

pliablepixels

---

## 🤖 Claude

Perfect! The build succeeds. Now let's commit this debug version so you can test it on iOS and see what the logs tell us:

---

## 🤖 Claude

pliablepixels

## 👤 User

Do a detailed code review. Follow guidelines from @AGENTS.md and assign a grade

---

## 🤖 Claude

I'll conduct a comprehensive code review following the AGENTS.md guidelines. Let me start by organizing this task.

pliablepixels

### Problem 1: **Instructions are buried in narrative text**
Your current CLAUDE.md has rules mixed with explanatory text:
```markdown
Global Rules
-------------
- All text needs to be internationalized. Always make sure all language files are in sync and updated with each change
```

This is easy to **read** but easy to **forget during execution**.

pliablepixels

import firebase_admin
from firebase_admin import messaging
from firebase_admin import exceptions
from flask import jsonify
import json
from firebase_admin import auth 
from functools import wraps
import jwt
import re

pliablepixels

use_sequence = yes

ml_sequence= {
                'general': {
                        'model_sequence': 'object,face,alpr'

                },
                'object': {
                        'general':{
                                'pattern':'(person)',

pliablepixels

Getting dlandon's docker image to work reliably with a GPU using a nvidia's pre-built CUDA/CuDNN image

Raw notes on what I did to get GPU/CUDA/CuDNN working in dlandon's docker image. Not optimized. So I don't forget later.

There are two ways:

1. Use his docker image which derives from phusion and manually install CUDA/cuDNN. This did not work for me - I got all sorts of segfaults, install errors, driver mismatch errors (even though my host/docker cuda versions were identical). When things worked after various pacakge errors installing cuda, the image failed after a restart. I've come to the conclusion that CuDNN/CUDA install is very sensitive to environments and what works for one person may dramatically fail for the other especially when we are trying to link a host GPU to a docker container. May as well leave this to the experts (nvidia). So the next option:

pliablepixels

import pyzm.api as zmapi

api_options = {
    'apiurl': 'https://yourserver/zm/api',
    'portalurl': 'https://yourserver/zm',
    'user': 'admin',
    'password': 'password',
    'logger': None # use none if you don't want to log to ZM
}

pliablepixels

dbell HTTP hook
===============

dbell HTTP hook usage:
User enter their server parameters, which will be stored in FTP field in http//example.com format, char HttpServer[256];
http server URL address example usage: http://192.168.1.183 where http server IP address is 192.168.1.183
 
Event:
dbell will do an HTTP POST to the server URL anytime someone presses the call button
dbell will do an HTTP POST to the server URL anytime a motion is detected