theodcr

#!/usr/bin/env python3

import os

from jupytext import cli

NOTEBOOK_DIR = 'jupyter'
FORMAT = 'py:percent'

ssophwang

{"bagpipe": {"on": 0.07142857142857142, "said": 0.07142857142857142, "Shall": 0.07142857142857142, "Did": 0.07142857142857142, "Ah": 0.07142857142857142, "didn": 0.35714285714285715, "I": 0.07142857142857142, "won": 0.07142857142857142, "lying": 0.07142857142857142, "if": 0.07142857142857142}, "pardon": {"the": 1.0}, "Gnu": {"murnood": 1.0}, "Night": {"s": 0.6666666666666666, "\n": 0.3333333333333333}, "yellow": {"caterpillar": 1.0}, "four": {"nickels": 0.2, "\n": 0.4, "is": 0.2, "more": 0.2}, "hath": {"spoken": 0.3333333333333333, "sent": 0.3333333333333333, "lent": 0.3333333333333333}, "Wretch": {"I": 1.0}, "sleep": {"late": 0.2, "\n": 0.6, "in": 0.2}, "SARDINES": {"\n": 1.0}, "Zoomed": {"up": 1.0}, "evermore": {"\n": 1.0}, "hate": {"\n": 1.0}, "Until": {"I": 0.3333333333333333, "a": 0.3333333333333333, "he": 0.3333333333333333}, "drinkin": {"that": 1.0}, "whose": {"foot": 0.3333333333333333, "velvet": 0.3333333333333333, "fiery": 0.3333333333333333}, "swam": {"away": 1.0}, "Claus": {"\n": 1.0}, "sweetest":

huihut

1. 以管理员身份运行CMD
2. slmgr.vbs /upk
3. slmgr /ipk W269N-WFGWX-YVC9B-4J6C9-T83GX
4. slmgr /skms zh.us.to
5. slmgr /ato

Krishna-kg732

Teaching Kubeflow to Talk to AI Agents

My GSoC 2026 with Kubeflow: Katib (Optimizer) Client Support for the Kubeflow MCP Server

Contributor: Krishna Gupta (@Krishna-kg732) · Org: Kubeflow (CNCF) Mentors: Abhijeet Dhumal, Akash Jaiswal, Dhanisha Phadate Project: Katib (HPO) Client Support for kubeflow-mcp-server · Medium · May 25 to August 24, 2026

---

I spent this summer teaching Kubeflow's hyperparameter tuning stack how to talk to AI agents.

ben-vargas

Amp Orbs Running CLIProxyAPI + Amp Router Custom URL

This guide runs CLIProxyAPI in an Amp orb around the clock and publishes it at a stable public HTTPS URL through Tailscale Funnel. An Amp Custom URL model-routing connection then sends Claude requests there, and CLIProxyAPI serves them with your Claude Pro or Max subscription (OAuth login) instead of Amp credits.

The host scripts, templates, and an agent skill that walks you through every step live in ben-vargas/amp-plugins/skills/amp-cli-proxy-api. The easiest path is to install the skill and ask Amp to set it up:

amp skill add --global ben-vargas/amp-plugins/skills/amp-cli-proxy-api

Maciejdziuba

The Software Factory Playbook — Dex Horthy's 4-gate workflow as a skill

A Claude Code Agent Skill built from Dex Horthy's (HumanLayer) playbook on David Ondrej's podcast.

"Once the model has written thousands of lines of code, it is harder to change. The sessions that generate design docs are context-light — you get the most model intelligence when you do the hard thinking early."

By default, agents build horizontally: all the backend, then all the frontend, then a 2,000-line diff lands in your lap and reviewing it is your problem. This skill flips that. Every decision that matters gets made before the code exists — where changing your mind costs a sentence, not a rewrite.

What it does

ebritanica

$kad['Санкт-Петербург'] = array([59.908456,29.659262],[59.885888,29.679213],[59.869938,29.74622],[59.863651,29.793783],[59.826761,29.824958],[59.812481,29.891431],[59.822211,29.967316],[59.816563,30.012207],[59.815913,30.076178],[59.801366,30.148366],[59.802641,30.173356],[59.812262,30.198768],[59.825935,30.233399],[59.833165,30.281853],[59.812214,30.319813],[59.811782,30.343651],[59.816016,30.380928],[59.825793,30.435141],[59.845107,30.458015],[59.853578,30.487691],[59.865128,30.527404],[59.889652,30.524301],[59.91036,30.526205],[59.927852,30.533823],[59.949187,30.543342],[59.966752,30.552995],[59.979217,30.525869],[59.984258,30.496359],[59.999108,30.47677],[60.012867,30.471708],[60.025186,30.450557],[60.041604,30.437535],[60.052483,30.404479],[60.066082,30.38521],[60.090189,30.371481],[60.095391,30.323246],[60.099381,30.278742],[60.092028,30.24515],[60.084021,30.214452],[60.07635,30.182448],[60.058877,30.144609],[60.045573,30.044505],[60.039066,29.980322],[60.028953,29.865704],[60.020797,29.741338],[60.0011

markasoftware

############ If you are using DOCKER all-in-one image, create Dockerfile like:         ################
############ FROM openproject/openproject:16                                           ################
############ COPY ./enterprise_token.rb app/models/enterprise_token.rb                 ################

############ If you are runing a manual installation:                                  ################
############ REPLACE app/models/enterprise_token.rb in the source code with this file! ################
############ also be sure to RESTART OpenProject after replacing the file.             ################

############ If using some other set up (eg docker-compose), read the comments on      ################
############ https://gist.github.com/markasoftware/f5b2e55a2c2e3abb1f9eefcdf0bfff45    ################

deonna

The prompt that removes 15 AI design tells

Paste this into Claude Code, Cursor, Codex, Lovable, Bolt, or whatever you build with. It has your agent install Impeccable (free and open source) and run it, then fix the 15 tells.

First, set up Impeccable, a free, open-source design skill for AI coding agents (github.com/pbakaus/impeccable, Apache 2.0). If you can run terminal commands:

1. From the project root, run: npx impeccable install --scope=project
   If it asks which tools to install for, pick the one you're running in.
2. Run /impeccable init so it records who the site is for and how it should feel. If the /impeccable command isn't available until I restart you, tell me and keep going with step 3.