devio

Draft and publish DevelopersIO articles end-to-end. Search existing tags to stay consistent with the site’s taxonomy, draft in English or Japanese, then publish to Contentful with tag resolution and optional auto-translation.

Install with your preferred package manager.

gh

gh skill install oharu121/skills devio

npx

npx skills add oharu121/skills/devio

pnpm

pnpm dlx skills add oharu121/skills/devio

What it does

Writing for DevelopersIO has two friction points: the Contentful API dance (tag IDs, content types, CMA tokens) and keeping article tags consistent with what’s already on the site. This skill handles both.

/devio article <topic> drafts a new article. It searches existing DevelopersIO tags so your draft uses the same taxonomy already in use (no “AWS” vs “Amazon Web Services” drift). /devio publish <file> takes a finished markdown file and pushes it to Contentful — resolving tag IDs, picking the right content type, and optionally translating the article to Japanese in a second entry.

The skill never reads .env directly. All Contentful API calls route through a contentful.py helper that reads the CMA token internally, so the secret never enters Claude’s context.

When to use it

• You’re a bilingual writer shipping to DevelopersIO and want the same article in English and Japanese without hand-syncing
• You want consistent tagging — the skill surfaces existing tags so you don’t create near-duplicates
• You prefer drafting in markdown and publishing via CLI over the Contentful web editor

When not to use it

This skill is shaped for DevelopersIO’s Contentful space specifically — content types, tag structure, and translation fields. Adapting it to a generic Contentful blog requires editing contentful.py and the article/publish subcommand files.

How it works

1. Draft with tag context. /devio article <topic> searches existing DevelopersIO tags matching the topic, shows you a shortlist, and drafts the article using the tags you pick. The result is a markdown file under drafts/.

2. Review locally. The draft is plain markdown. Edit it in your usual tools — the skill doesn’t own the writing surface.

3. Publish to Contentful. /devio publish drafts/<file>.md resolves tag names to Contentful IDs, picks the content type, and creates the entry. Asks whether to auto-translate to Japanese as a linked entry.

4. Verify in Contentful. The skill returns the entry URL. Publish status is draft until you flip it in Contentful’s UI — the skill doesn’t auto-publish, by design.

Example

/devio article Setting up Bedrock KB with S3

The skill searches for existing tags like AWS, Bedrock, RAG, shows you the matches, and drafts the article with those tags pre-attached. You edit drafts/bedrock-kb-s3-setup.md locally, then:

/devio publish drafts/bedrock-kb-s3-setup.md

It resolves the tags, creates the Contentful entry, and asks “Translate to Japanese?” — yes creates a linked Japanese entry. Both remain drafts; you flip them live from Contentful’s UI.

Security model

• .env is never read directly by the skill
• All Contentful API calls go through scripts/contentful.py, which reads the CMA token internally
• The CMA token never enters Claude’s context
• Output the token or .env contents is not possible through the skill’s interface

Related skills

generate-bruno-files When writing API articles, scaffold the sample .bru files alongside.

release For article series published as versioned bundles, use release automation.

Source

View SKILL.md on GitHub — article and publish subcommands, Contentful schema, translation workflow.