- Home
- Skills
- Technical Documentation
- blog style ghostwriter
blog style ghostwriter
Drafts Jekyll-ready engineering blog posts from technical notes, code snippets, and fragmented outlines.
Free
Works with the AI Agents you already use
blog style ghostwriter
Example session with this skill installed
I have these notes on implementing a basic Raft consensus algorithm in Go. It's for my tech blog. Make a Jekyll draft with frontmatter. Focus on leader election.
- Read your context and instructions
- Compiled the blog style ghostwriter
I have converted your notes on Raft leader election into a structured Jekyll blog post, including a Table of Contents and full frontmatter.
Jekyll Post Draft
---
layout: post
title: "Understanding Raft: Implementing Leader Election in Go"
description: "A deep dive into the mechanisms behind leader election in the Raft consensus protocol, featuring a reference implementation in Go."
date: 2023-10-27
feature_image: /images/raft-election-thumb.png
tags: [distributed-systems, go, consensus]
---
Raft is a consensus algorithm designed to be easy to understand while remaining functionally equivalent to Paxos. This post explores the leader election phase, explaining how nodes transition from followers to candidates and eventually to leaders.
<!--more-->
## Table of Contents
- [The Three States](#the-three-states)
- [Election Timeouts](#election-timeouts)
- [Implementation in Go](#implementation-in-go)
- [Conclusion](#conclusion)
## The Three States
In a Raft cluster, every node exists in one of three states: Follower, Candidate, or Leader. Nodes start as followers. If they receive no communication from a leader, they become candidates and request votes from the rest of the cluster.
---
## Election Timeouts
To prevent split votes where multiple nodes become candidates simultaneously, Raft uses randomized election timeouts. Typically, these range from 150ms to 300ms.
{% include image_caption.html imageurl="/images/raft-timeout-diagram.png" title="Raft Election Timeout Flow" caption="Randomized timers prevent concurrent candidate transitions." %}
## Implementation in Go
The following snippet demonstrates the state transition logic when a heartbeat is missed:
```go
func (rf *Raft) attemptElection() {
rf.mu.Lock()
rf.state = Candidate
rf.currentTerm++
rf.votedFor = rf.me
rf.mu.Unlock()
// Request votes from peers...
}
Conclusion
Randomized timeouts are the cornerstone of Raft's simplicity. By ensuring that one node is likely to time out before others, the protocol maintains a stable leader without complex negotiation.
Related Posts / Websites 👇
**Next steps**
- Replace the `/images/` placeholders with your actual diagram files.
- Verify the `date` in the frontmatter matches your intended publishing schedule.
- Add specific log output examples to the Implementation section.
Connects securely to your tools. The creator never sees your data.
About this skill
Writing technical blog posts often stalls at the gap between raw engineering notes and a polished, publishable draft. This skill bridges that gap by transforming fragmented ideas, code snippets, and screenshots into structured Jekyll-ready posts that maintain a consistent technical voice.
What it does
- Structure Extraction converts disorganized notes into logical section hierarchies with specific reader promises.
- Jekyll Scaffolding generates full frontmatter including layouts, tags, and feature image paths.
- Tone Mimicry applies a calm, lecture-like technical voice that moves from concept to mechanism.
- Bilingual Support produces EN/ZH content where the English explanation precedes the Chinese translation.
- Liquid Integration handles image placement using repository-specific
image_caption.htmlpatterns.
How it works
- Identify type by determining if the material is a technical explainer, setup log, or lecture note.
- Draft skeleton including table of contents and internal links based on the provided material.
- Expand prose using an house-style profile to fill gaps between technical evidence and code blocks.
- Finalize metadata by applying specific Jekyll frontmatter and Jekyll-compatible Liquid includes.
Frameworks & tools
Specifically designed for Jekyll-based engineering blogs. It utilizes Liquid templates for image handling and follows standard Markdown conventions for technical documentation, including MathJax/LaTeX formulas and code fences.
Why this beats prompting it yourself
Generic AI writing often sounds like marketing fluff or diary entries. This skill enforces a specific engineering-first hierarchy, ensures Table of Contents consistency, and prevents the fabrication of fake technical benchmarks or API details.
Use cases
- Drafting from fragments by turning a rough outline and three code blocks into a full tutorial.
- Polishing technical logs to remove personal diary-style reflections and align with a formal blog voice.
- Bilingual documentation for teams needing technical explainers in both English and Chinese.
- Image-heavy tutorials where screenshots need specific Liquid caption wrappers and descriptive titles.
Known limitations
Does not support generating actual image files or verifying external link health. Requires a style profile reference to perfectly match an existing repository voice.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Technical Documentation.
Cannot generate image assets. Requires manual style profile input for non-standard voices. Does not verify external API validity or real-world benchmarks.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
1 use
Times used to date
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 27 days ago
- Free with an Agensi account
- Software
Python