agentleFS
fertile-orgfertile-org/coppice/.agents/skills/blog-writing/SKILL.md

blog-writing

Skill27 starsChanged 7 days agoApache-2.0View on GitHub

Use when planning, drafting, reviewing, or rewriting a post for the Coppice blog (website/src/content/blog/). Covers what a post is for, how to pick the topic and the reader's real problem, the structure, how much Coppice detail to include, and the checks to run. Not for user docs, READMEs, or social posts.

10 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

10 rules. 0 are common, 4 are distinctive, and 6 are very specific to this project.

  • 0

    Common

    Most of the 5 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 6

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Start from a generic question, not a Coppice feature: "How to use Claude Code and Codex together on one project", not "Assigning agents in …

    line 14
  2. Very specific:

    Before you write, state the problem the reader actually has.

    line 20
  3. Distinctive:

    If you are not sure what the real problem is, ask the owner before you draft.

    line 24
  4. Distinctive:

    Keep the whole post short.

    line 34
  5. Very specific:

    Write for someone who has never opened Coppice.

    line 38
  6. Very specific:

    Leave out internal terms and settings: connectors, auto_start_runs, column rules, exact error messages, every button on a form.

    line 40
  7. Distinctive:

    Leave out edge cases and troubleshooting.

    line 41
  8. Very specific:

    Do not recommend or pitch a competing product.

    line 47
  9. Distinctive:

    Use one or two at most, only where a picture saves words.

    line 51
  10. Very specific:

    Delete any image the post no longer uses from website/public/blog/<slug>/.

    line 53

“Common” is measured: kept by at least 6 in 10 of the 5 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Blog writing

The Coppice blog exists for search. Each post answers a question people already type into Google about coding agents, then shows Coppice as one good way to solve it. A post is not a getting-started guide and not product documentation.

For English prose, also follow english-technical-writing. For frontmatter, routes, images, and checks, see website/BLOG.md.

1. Pick a topic people search for

  • Start from a generic question, not a Coppice feature: "How to use Claude Code and Codex together on one project", not "Assigning agents in Coppice".
  • Phrase the title the way someone would search for it. "How to …" and "… vs …" titles work well.
  • One post answers one question.

2. Name the reader's real problem

Before you write, state the problem the reader actually has. Do not invent one to set up Coppice.

Check it against how people work today. Example: for using Claude Code and Codex together, the hard part is not telling who changed what. One agent can plan and another can build from that plan without trouble. The hard part is switching between terminals, and it gets worse with several tasks in parallel.

If you are not sure what the real problem is, ask the owner before you draft.

3. Structure

  1. answer (frontmatter): the direct answer in two to four sentences. It mentions the options, including Coppice.
  2. The problem: a short section on why this is hard, in the reader's terms.
  3. The options: two to four ways to solve it. Most should work without Coppice, such as plain terminals, git worktrees, or a script. Give each one a sentence or two on when it works and where it stops working. Show a command only when it is the whole option, such as git worktree add.
  4. The Coppice option: what Coppice does for this problem in one or two sentences, then at most five short steps to try it. Link to the docs for the details.
  5. Which one to pick: one line per option, matched to a situation.

Keep the whole post short. A reader should finish it in a few minutes.

4. How much Coppice detail

Write for someone who has never opened Coppice.

  • Leave out internal terms and settings: connectors, auto_start_runs, column rules, exact error messages, every button on a form. Say "pick Claude Code or Codex for each agent", not "choose a connector".
  • Leave out edge cases and troubleshooting. Link to Getting started, Concepts, or Providers instead.
  • Each Coppice claim must match the current docs or code. Check website/src/pages/docs/ and, when in doubt, the server code. Do not promise a feature that is not shipped.

5. Other tools

  • Generic approaches are fine to describe: terminals, git worktrees, shell scripts, the CLIs themselves.
  • Do not recommend or pitch a competing product. If one is relevant, mention it in a few words at most and move on to Coppice.

6. Screenshots

  • Use one or two at most, only where a picture saves words.
  • Captions describe what the reader sees. Do not write "Demo data" or "fixture" in a caption.
  • Delete any image the post no longer uses from website/public/blog/<slug>/.

7. Before you finish

From website/:

npm run build
npm run check:seo
npm run check:links
node --experimental-strip-types --test src/lib/blog.test.ts

Then read the post once as a new visitor who searched for its title. Ask:

  • Does the first paragraph answer the question?
  • Is the problem one the reader would recognize?
  • Could they try the Coppice option from the steps alone, and find more in the linked docs?
  • Is there any sentence that only makes sense to someone who already uses Coppice? Cut it.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.