---
title: "George Sims (@george-sims) | Hashnode"
description: "Senior DevOps Engineer @ Bynder. Read the latest articles by George Sims on Hashnode."
url: "https://hashnode.com/@george-sims"
---

@george-sims

# George Sims

@george-sims · Barcelona, Spain · Joined February 2026

Senior DevOps Engineer @ Bynder

### [Why a Bigger Context Window Makes Claude Worse](https://downtherabithole.dev/why-a-bigger-context-window-makes-claude-worse?utm_source=hashnode&utm_medium=feed)

Research shows model recall drops as context grows, even on simple tasks, and security monitors miss more attacks in longer transcripts.

10 min read · [downtherabithole.dev](https://downtherabithole.dev/why-a-bigger-context-window-makes-claude-worse?utm_source=hashnode&utm_medium=feed)

[Aug 27](https://hashnode.com/posts/why-a-bigger-context-window-makes-claude-worse/6a8fdf441904f27dcb7d8e9b)

### [Looking Through Claude's Context Window](https://downtherabithole.dev/looking-through-claude-s-context-window?utm_source=hashnode&utm_medium=feed)

What actually fills a Claude Code context window before you type anything, how memory, tools, skills and MCP differ, and why sub-agents are important

9 min read · [downtherabithole.dev](https://downtherabithole.dev/looking-through-claude-s-context-window?utm_source=hashnode&utm_medium=feed)

[4](https://hashnode.com/posts/looking-through-claude-s-context-window/6a833d8ec3b7bb972b00d568) [Aug 17](https://hashnode.com/posts/looking-through-claude-s-context-window/6a833d8ec3b7bb972b00d568)

### [What Is a Token?](https://downtherabithole.dev/what-is-a-token?utm_source=hashnode&utm_medium=feed)

How LLM tokenization actually works, with a real worked example, and why it's the exact unit your Claude Code bill is calculated in.

5 min read · [downtherabithole.dev](https://downtherabithole.dev/what-is-a-token?utm_source=hashnode&utm_medium=feed)

[Aug 16](https://hashnode.com/posts/what-is-a-token/6a81de366d6f13c5be271bf9)

### [Distributed Tracing for CI Pipelines Without Touching a Single Workflow File](https://downtherabithole.dev/distributed-tracing-for-ci-pipelines-without-touching-a-single-workflow-file?utm_source=hashnode&utm_medium=feed)

How to get org-wide distributed tracing for GitHub Actions with OTEL's githubreceiver and a single webhook, plus a method for sizing and justifying.

8 min read · [downtherabithole.dev](https://downtherabithole.dev/distributed-tracing-for-ci-pipelines-without-touching-a-single-workflow-file?utm_source=hashnode&utm_medium=feed)

[3](https://hashnode.com/posts/distributed-tracing-for-ci-pipelines-without-touching-a-single-workflow-file/6a771f53e42cf22dee8376a1) [Aug 8](https://hashnode.com/posts/distributed-tracing-for-ci-pipelines-without-touching-a-single-workflow-file/6a771f53e42cf22dee8376a1)

### [Migrating A Critical Kubernetes Deployment From The Default Namespace Without Any Downtime](https://downtherabithole.dev/migrating-a-critical-kubernetes-deployment-from-the-default-namespace-without-any-downtime?utm_source=hashnode&utm_medium=feed)

Production case study: moving a critical Kubernetes service to a new namespace with zero downtime using ExternalName proxying

7 min read · [downtherabithole.dev](https://downtherabithole.dev/migrating-a-critical-kubernetes-deployment-from-the-default-namespace-without-any-downtime?utm_source=hashnode&utm_medium=feed)

[Aug 2](https://hashnode.com/posts/migrating-a-critical-kubernetes-deployment-from-the-default-namespace-without-any-downtime/6a6f3090d0571c4ac05cfad8)
