rossant

AI Agent Repo Bootstrap

Human usage

Copy this file into the root of a software repository.

Then tell your coding agent:

Read AGENT_BOOTSTRAP.md and bootstrap this repo.

rossant

Local Continuous Recording “Black Box” System

1. Problem Statement

There is currently no simple, affordable, and reliable system that allows continuous, long-term indoor video recording (6–12 months) with the following properties:

• Full 24/7 recording without gaps
• Local-only operation (no cloud, no external connectivity)
• User-controlled storage
• Simple deployment (plug-and-play)

rossant

# SPDX-License-Identifier: MIT
#
# Copyright (c) 2025 Cyrille Rossant
#
# Permission is hereby granted, free of charge, to any person obtaining a copy
# of this software and associated documentation files (the "Software"), to deal
# in the Software without restriction, including without limitation the rights
# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
# copies of the Software, and to permit persons to whom the Software is
# furnished to do so, subject to the following conditions:

rossant

import requests
import pandas as pd

def get_author_id(author_name):
    """Retrieve OpenAlex author ID using the author's name."""
    search_url = f"https://api.openalex.org/authors?search={author_name}"
    response = requests.get(search_url)
    response.raise_for_status()
    data = response.json()
    if data['results']:

rossant

// Requires gcc, the Vulkan SDK, and libglfw3-dev
// On Linux, compile with:
// gcc -Wall -o fail fail.c -Ibuild/_deps/glfw-src/include/ -lglfw -lvulkan

#include <assert.h>
#include <stdbool.h>
#include <string.h>
#include <vulkan/vulkan.h>

#include <GLFW/glfw3.h>

rossant

def lru_cache(maxsize=0):
    """Custom LRU cache implementation that gives access to the underlying cache dictionary. [better to used functools one if you don't need this]"""
    def wrap(f):
        cache = {}
        last_used = []

        def wrapped(*args):
            if args in cache:
                # HIT.
                # Update the last_used list.

rossant

from onelib import one

"""
Different ONE backends are available, HTTP, figshare, etc.
One has to implement the following functions to create a new ONE backend:

* list_all_files(): return a list of relative file paths for ALL available files.

* search(...): return a list of dset_ids (by default, a dset_id is a relative file path).
  The default implementation calls list_all_files(), and performs the search directly on that list.

rossant

"""Write a NumPy array in parallel from multiple CPUs/processes, using shared memory."""

from contextlib import closing
import multiprocessing as mp
import os

import numpy as np


def _init(shared_arr_):

rossant

import os
import requests
import shutil


def _dl(url, path):
    print("download", url, "to", path)
    response = requests.get(url, stream=True)
    if response.status_code != 200:
        return

rossant

#!/usr/bin/env python3

"""

Code from http://www.labri.fr/perso/nrougier/python-opengl/#the-hard-way

"""

import ctypes
import logging