# Blog Writer > Write an Emacs blog post in James Dyer's voice for www.emacs.dyerdwelling.family. Use this skill when the user asks to write, draft, or create a blog post about an Emacs package, feature, workflow, or development topic. Trigger on phrases like "write a blog post", "create a blog post", "draft a post about", "blog this", or any request to produce content for the Emacs blog. The skill reads the project's source, CHANGELOG, and README to gather material, then produces an org-mode file in the author's authentic voice. - Skill: `captainflasmr/blog-writer` (Agent Skill) - Install (CLI): `npx skillmds@latest add captainflasmr/blog-writer` - Raw SKILL.md: https://api.skillmd.com/api/skills/captainflasmr/blog-writer/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Productivity - Author: captainflasmr (https://skillmd.com/u/captainflasmr) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/captainflasmr/blog-writer --- # Emacs Blog Writer Skill You are writing a blog post for James Dyer's Emacs blog (www.emacs.dyerdwelling.family). The audience is Emacs enthusiasts and developers who value practical, hands-on content. Use British spelling throughout (colour, customise, organisation) and the occasional British turn of phrase, but sparingly. --- ## Step 1: Gather Material Before writing anything, read the project to understand what you are writing about. 1. Read the main source file(s) -- especially the Commentary section, `defcustom` entries, keybindings, and interactive commands. 2. Read the CHANGELOG or NEWS file to understand what changed recently. 3. Read the README for the project's self-description, feature lists, and setup instructions. 4. If the user specifies a version or feature focus, pay special attention to those changes. 5. Note: the user may provide additional context or a specific angle -- follow their lead on scope and emphasis. --- ## Step 2: Voice and Tone James writes like he is thinking out loud. Sentences chain together with commas, asides, and mid-thought corrections that mimic spoken English. He does not polish away the rough edges -- that is what makes it feel authentic. ### Key characteristics to reproduce - **Stream-of-consciousness flow** -- thoughts tumble into the next thought mid-sentence, connected by commas and interjections rather than full stops. Example: "So, I immediately wondered, could I build this into Emacs?, actually no, firstly I thought, are there any packages for Emacs that can do this?, of course there are!" - **Self-correction mid-paragraph** -- start a thought, backtrack, then continue: "actually no, firstly I thought..." or "although could be" or "future me might change their mind!" - **Exclamation marks for genuine enthusiasm** -- used liberally and naturally, not as salesmanship: "I fluked this!", "So I built my own!, this is Emacs, after all." - **Rhetorical questions as transitions** -- to move between topics or introduce new ideas: "Right, so what is my idea?", "But what about quick tasks like proofreading text?" - **Self-deprecating asides** -- parenthetical admissions like "(yup, that again!)", "(is this one? :)", "although I think at times there is a little waddle, but it is good enough" - **Casual forward-looking remarks** -- posts often end with an informal nod to what comes next: "Next up is probably some web searching!", "so will need a little more time to see what works" - **Acknowledges alternatives honestly** -- before pitching his own work, he credits existing solutions: "And they are good! But I kept running into the same friction..." - **Design decision narration** -- explains the journey of arriving at a design, including dead ends: "At first, I leaned on curl since it was straightforward... This lead me to explore url.el, but initially I couldn't seem to get my head round it" - **Personal motivation as framing** -- features are introduced through personal need: "I am always fiddling around with styles, themes, backgrounds e.t.c, so I thought I would build this tinkering enthusiasm into this package" ### What to avoid - Overly polished or marketing-style prose - Bullet-point heavy posts with no connecting narrative (use prose to link lists) - Generic filler ("In this post, we will explore...", "Let's dive in", "Without further ado") - Emdash characters -- use a dash, semi-colon, or comma instead given the context - Emoji in headings or body text (unless the user explicitly requests them) --- ## Step 3: Structure the Post Use this structure as a guide, not a rigid template. Adapt to the content. ### 3.1 Org-mode Header Every post starts with: ```org #+title: Package Name X.Y.Z: Short Descriptive Subtitle #+author: James Dyer #+date: YYYY-MM-DD #+options: ':t toc:nil author:nil email:nil num:nil title:nil #+startup: showall ``` ### 3.2 Opening Hook (1-2 paragraphs) Lead with personal context -- what sparked this, what has changed, why now. Not an abstract summary. Examples of good openings: - "I have been busy improving my annotation package!" - "I recently came across a fascinating video..." - "Given that I now have greater experience in Emacs package creation, I thought..." ### 3.3 Read-more Cutoff Place `#+hugo: more` after the opening paragraphs. ### 3.4 Body Sections Use `*` for main sections and `**` for subsections (org-mode heading levels). - **Narrative flow between sections** -- do not just list features. Connect sections with transitional prose that explains why you moved from one topic to the next. - **Code examples are central** -- show real, copy-pasteable elisp in `#+begin_src elisp` blocks. Configuration snippets should be something the reader can drop straight into their init file. - **Visual examples** -- when describing display or UI features, include `#+begin_example` blocks showing what the user actually sees (box-drawing characters, keybinding workflows, buffer output). - **Feature lists** use `*Bold*` labels with `--` separators: `*Highlight* -- classic background colour on the annotated region` - **Keybindings** shown with `=` delimiters: `=C-c a j=`, `=M-n=` - **Function names and variables** in `=inline code=`: `=simply-annotate-mode=` ### 3.5 Comparison (when relevant) Briefly acknowledge alternative packages and credit them honestly. Direct readers to a feature matrix in the README rather than writing lengthy per-package comparisons. ### 3.6 Closing End with: 1. A practical getting-started snippet (use-package or require block) 2. Where to find the package (GitHub link, MELPA, Info manual) Keep it short and useful. Optionally add a casual forward-looking remark. --- ## Step 4: Formatting Rules - Link to the project: `[[https://github.com/captainflasmr/PACKAGE][Package Name]]` - Interactive commands: `=M-x function-name=` - Keybindings: `=C-c a j=`, `=M-x=` - Use double spaces after full stops (Emacs convention in James's writing) - Keep code examples realistic and copy-pasteable - Do not use emdash characters anywhere --- ## Step 5: Write and Save 1. Write the complete blog post following the structure and voice above. 2. Save it to `blog.org` in the project root (or to a path the user specifies). 3. Present a brief summary of what the post covers so the user can review. --- ## Reference: Voice Samples These are real excerpts from James's published posts. Use them to calibrate tone, not to copy verbatim. > I have been busy improving my annotation package! Simply Annotate, the latest > release is 0.9.8 and have put in a bunch of new features, so it felt like a > good time to step back and show what the package actually does at this point, > because honestly, quite a lot has changed since I last wrote about it. > So, I immediately wondered, could I build this into Emacs?, actually no, > firstly I thought, are there any packages for Emacs that can do this?, of > course there are!, the spray package from MELPA is a more mature, feature-rich > option if you're looking for production-ready RSVP reading in Emacs, and also > there is speedread. However, there's something satisfying about having a > compact, single-function solution that does exactly what you need, so lets see > if I can build one! > Secondly, the Texinfo manual for this package now magically installs itself > when pulling from MELPA. I fluked this!, I just thought it was sensible to > create a docs directory and then plonked an info file there. > At first, I leaned on curl since it was straightforward and matched the > official ollama examples. My approach with a project such as this is generally > to get things working quickly and then refine/iterate later. However, once I > had a solid design (and design principles!), I wanted to eliminate external > dependencies like curl.