GitHub lets you suffix any PR with .diff or .patch to get the raw diff/patch.
e.g.
https://github.com/github/gitignore/pull/4908.diff
https://github.com/github/gitignore/pull/4908.patch
| { | |
| "env": { | |
| "ANTHROPIC_BASE_URL": "http://127.0.0.1:1234", | |
| "ANTHROPIC_AUTH_TOKEN": "dummy", | |
| "ANTHROPIC_MODEL": "google/gemma-4-e4b" | |
| }, |
| # https://gist.github.com/ElijahLynn/adf761215c85a5d8136ced38d4022ddd | |
| function lid-sleep --description "Toggle or check lid-close sleep behaviour" | |
| set config /etc/systemd/logind.conf.d/lid.conf | |
| switch $argv[1] | |
| case help --help -h | |
| echo "Usage: lid-sleep [on|off|status|help]" | |
| echo "" | |
| echo " on Re-enable sleep when the lid is closed" | |
| echo " off Prevent sleep when the lid is closed" |
| #!/usr/bin/env python3 | |
| """ | |
| Whole-GPU utilization sampler via the i915 PMU -- the same source btop and | |
| intel_gpu_top use. Reads per-engine `*-busy` perf counters (nanoseconds the | |
| engine was busy) and normalizes by PERF_FORMAT_TOTAL_TIME_ENABLED: | |
| engine_util% = delta(busy_ns) / delta(time_enabled_ns) * 100 | |
| This measures the GPU system-wide (pid=-1), so it includes warp's rendering | |
| AND the GNOME compositor compositing warp's frames -- matching btop. |
| #!/bin/bash | |
| # Script to "kick" Sony XM6 headphones when multi-point switching gets stuck | |
| CARD="bluez_card.80_99_E7_DE_5F_68" | |
| echo "Attempting to reset Bluetooth audio for Eli Beats..." | |
| # 1. Toggle profile to "off" and back to "a2dp-sink" | |
| # This usually forces the transport to re-handshake without a full disconnect | |
| echo "-> Resetting profile..." |
Notes for fixing extension activation failures and Python environment warnings when using Ansible 26.6+, Python Environments, and ms-python.python in Cursor on macOS.
Tested on Cursor ~1.105 (VS Code 1.105 base), macOS arm64, June 2026.
ms-python.pythonTesting the Claude Code agentic loop (claude -p --dangerously-skip-permissions) building
a small Python CLI from scratch, with two effort levels compared.
random, list, add, delete subcommands, stdlib onlywhile grep -q '\- \[ \]' implementation_plan.md — one task per iteration, auto-terminatesmain (xhigh/default effort) vs no-thinking (--effort low)| dconf reset -f /org/gnome/desktop/peripherals/tablets/056a:52b5/ | |
| echo i2c-WACF2200:00 | sudo tee /sys/bus/i2c/drivers/i2c_hid_acpi/unbind | |
| echo i2c-WACF2200:00 | sudo tee /sys/bus/i2c/drivers/i2c_hid_acpi/bind |