---
title: "Khasky (@khasky) | Hashnode"
description: "Developer with founder delusions. Read the latest articles by Khasky on Hashnode."
url: "https://hashnode.com/@khasky"
---

@khasky

# Khasky

@khasky · Joined August 2026

Developer with founder delusions

### [Practical AI Coding Workflows with Awesome Agent Skills](https://khasky.hashnode.dev/practical-ai-coding-workflows-with-awesome-agent-skills?utm_source=hashnode&utm_medium=feed)

An AI agent can finish a feature, pass the tests, and still leave behind unnecessary abstractions, misleading comments, or a README that no longer matches the product. Asking it to "check everything"

8 min read · [khasky.hashnode.dev](https://khasky.hashnode.dev/practical-ai-coding-workflows-with-awesome-agent-skills?utm_source=hashnode&utm_medium=feed)

[Oct 18](https://hashnode.com/posts/practical-ai-coding-workflows-with-awesome-agent-skills/6ad51564ceb820a098a70662)

### [Non-English Prompts Cost More Tokens, Context and Money in Most LLMs](https://khasky.hashnode.dev/non-english-prompts-cost-more-tokens-context-and-money-in-most-llms?utm_source=hashnode&utm_medium=feed)

Claude Opus 5 costs $5 per million input tokens whether a prompt arrives in English or in Russian. What changes between the two languages is how many tokens the same paragraph becomes, and the bill is

7 min read · [khasky.hashnode.dev](https://khasky.hashnode.dev/non-english-prompts-cost-more-tokens-context-and-money-in-most-llms?utm_source=hashnode&utm_medium=feed)

[2](https://hashnode.com/posts/non-english-prompts-cost-more-tokens-context-and-money-in-most-llms/6ac3290df7e50888e2792a2a) [Oct 5](https://hashnode.com/posts/non-english-prompts-cost-more-tokens-context-and-money-in-most-llms/6ac3290df7e50888e2792a2a)

### [Fable Becomes the Advisor to Opus 5.5 in Claude Code](https://khasky.hashnode.dev/fable-becomes-the-advisor-to-opus-5-5-in-claude-code?utm_source=hashnode&utm_medium=feed)

With Opus 5.5 as the main model in a Claude Code session, a Fable quota can still go to work without Fable taking over the session. The /advisor fable command connects Fable as an advisor that Opus co

3 min read · [khasky.hashnode.dev](https://khasky.hashnode.dev/fable-becomes-the-advisor-to-opus-5-5-in-claude-code?utm_source=hashnode&utm_medium=feed)

[Oct 3](https://hashnode.com/posts/fable-becomes-the-advisor-to-opus-5-5-in-claude-code/6ac056a397a6376f24f88546)

### [Rejudge Replaces Self-Review With 3 Independent Models and a Judge](https://khasky.hashnode.dev/rejudge-replaces-self-review-with-3-independent-models-and-a-judge?utm_source=hashnode&utm_medium=feed)

I have a hard time trusting an AI coding agent to review code it just wrote. The model already made the decision that the implementation was good enough to produce. Then we ask it to inspect that same

5 min read · [khasky.hashnode.dev](https://khasky.hashnode.dev/rejudge-replaces-self-review-with-3-independent-models-and-a-judge?utm_source=hashnode&utm_medium=feed)

[Oct 1](https://hashnode.com/posts/rejudge-replaces-self-review-with-3-independent-models-and-a-judge/6abed6497f021db32b3f6574)

### [Gemini 4 Argon Beats GPT-6 Astra and Claude on Most Benchmarks](https://khasky.hashnode.dev/gemini-4-argon-beats-gpt-6-astra-and-claude-on-most-benchmarks?utm_source=hashnode&utm_medium=feed)

Google DeepMind has announced Gemini 4 Argon, a frontier model aimed at long software engineering jobs, legal and finance work, and cyber defense. In Google's own evaluation it beats GPT-6 Astra, Clau

5 min read · [khasky.hashnode.dev](https://khasky.hashnode.dev/gemini-4-argon-beats-gpt-6-astra-and-claude-on-most-benchmarks?utm_source=hashnode&utm_medium=feed)

[Oct 1](https://hashnode.com/posts/gemini-4-argon-beats-gpt-6-astra-and-claude-on-most-benchmarks/6abdf4473aa34b18488b4043)