agentleFS
yohei1126yohei1126/agent-skills/devto-article/SKILL.md

devto-article

Skill2 starsChanged 22 days agoMITView on GitHub

Write, revise, or draft-register English technical articles for dev.to. Do not include API keys or bot-specific ops.

The file

Show its rules and text

We split this skill into the rules it gives an agent

21 rules. About 0 are common, 15 are distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈15

    Distinctive · estimate

    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. Distinctive:

    Write a new dev.to article

    line 11
  2. Distinctive:

    Check a draft before or after registering it as an unpublished draft via the API

    line 13
  3. Distinctive:

    Before drafting or revising, read:

    line 18
  4. Distinctive:

    When the same piece also goes to LinkedIn, keep claims and numbers aligned.

    line 22
  5. Very specific:

    When the piece already published elsewhere, set canonical_url to that first URL (for a LinkedIn cross-post, the LinkedIn Pulse URL)

    line 40
  6. Distinctive:

    Do not put version numbers in the title or URL slug

    line 46
  7. Very specific:

    Follow english-tech-blog-structure.

    line 51
  8. Distinctive:

    Use a cover image only when rights are clear

    line 57
  9. Distinctive:

    Wait a few days, then post on DEV

    line 65
  10. Distinctive:

    Keep claims, numbers, and how limits are stated aligned.

    line 67
  11. Very specific:

    If DEV ships first, set canonical_url on the later channel to the URL that published first.

    line 71
  12. Distinctive:

    Do this only when the owner asks to register a draft on DEV

    line 79
  13. Very specific:

    Create unpublished (draft) only.

    line 80
  14. Very specific:

    Before creating, check GET /api/articles/me/unpublished for an existing draft of the same piece

    line 83
  15. Very specific:

    After create, verify with GET /api/articles/me/unpublished, and confirm the article is not in published (GET /api/articles/me/published).

    line 84
  16. Distinctive:

    After registration, give the owner the draft URL or ID and leave publishing to them

    line 85
  17. Distinctive:

    Do not exaggerate. State limits

    line 89
  18. Distinctive:

    Use numbers only from measurement logs or public sources

    line 90
  19. Distinctive:

    Do not promise unconfirmed publish dates

    line 91
  20. Distinctive:

    Do not add consulting-style CTAs

    line 92
  21. Distinctive:

    Do not write as if you verified something you did not verify

    line 93

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “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

dev.to articles

Generic workflow for writing and revising English technical articles for dev.to. Put environment-specific ops—API key storage, repo paths, which model to use—in each assistant's shared rules, not here.

When to use

  • Write a new dev.to article
  • Adapt an English article from another channel (for example LinkedIn) for dev.to
  • Check a draft before or after registering it as an unpublished draft via the API

English writing rules

  1. Body language is English.
  2. Before drafting or revising, read:
    • english-tech-blog-structure — structure, headings, titles, closings, technical honesty
    • humanizer — strip sentence-level AI tells
    • llm-evaluation — when the draft cites eval or LLM measurement numbers
  3. When the same piece also goes to LinkedIn, keep claims and numbers aligned; change only channel constraints (headings, cover image, series UI).

DEV front matter

Example YAML at the top of the article, or in a separate meta file (see Cross-posting):

---
title: Specific claim or payoff
published: false
tags: neo4j, knowledgegraph, llm, cypher
canonical_url: https://www.linkedin.com/pulse/...
series: Ontographia
---
  • Drafts must keep published: false. Only the owner flips the flag or hits publish
  • tags are lowercase English, following site conventions. Aim for up to 4 tags for API registration. Forem/Dev.to create accepts at most 4 tags. If a fifth is desired for humans, note it in the file only; it cannot be sent on create
  • When the piece already published elsewhere, set canonical_url to that first URL (for a LinkedIn cross-post, the LinkedIn Pulse URL)
  • For a series, use series so it appears in DEV's series feature

Titles and slugs

  • Titles should state a concrete claim or payoff. Avoid vague launch phrasing
  • Do not put version numbers in the title or URL slug
  • For a series, Part N in the title is fine

Article shape

Follow english-tech-blog-structure. DEV-specific extras:

  • Open with the problem or conclusion in 2–3 sentences. Do not start with meta foreshadowing
  • For a series, place the series position (links to other parts) near the top
  • For a general audience, prefer Releases, docs, and the main repository over individual PRs
  • Markdown tables and fenced code are fine. Mermaid is environment-dependent, so also provide important diagrams as images or short text figures
  • Use a cover image only when rights are clear
  • Covers and diagrams: load visual-color and visual-diagram.

Cross-posting (LinkedIn → DEV)

Common order:

  1. Publish on LinkedIn
  2. Wait a few days, then post on DEV
  3. Set DEV canonical_url to the first-published LinkedIn Pulse URL
  4. Keep claims, numbers, and how limits are stated aligned; change only channel greetings and series UI

Series TOC links for published parts may keep the LinkedIn Pulse URLs.

If DEV ships first, set canonical_url on the later channel to the URL that published first.

Meta-only layout

For a LinkedIn → DEV cross-post, the LinkedIn article Markdown can remain the canonical body. Registration metadata may live separately (for example devto/<series>/YYYY-MM-DD-….yml) with title, published: false, tags, canonical_url, series, and source pointing at the LinkedIn path. Do not require a full copy of the article under devto/.

Drafting via the API

  • Do this only when the owner asks to register a draft on DEV
  • Create unpublished (draft) only. Do not call a publish API. Keep published: false
  • Send at most 4 tags. A fifth noted for humans stays out of the create payload
  • A repo may provide a script (example name scripts/devto_payload.py) that prints API JSON from the meta file and the LinkedIn source. The script does not call the API. Registration is a separate step, still with published: false
  • Before creating, check GET /api/articles/me/unpublished for an existing draft of the same piece
  • After create, verify with GET /api/articles/me/unpublished, and confirm the article is not in published (GET /api/articles/me/published). A public GET /api/articles/{id} can 404 while the article is unpublished
  • After registration, give the owner the draft URL or ID and leave publishing to them

Honesty

  • Do not exaggerate. State limits
  • Use numbers only from measurement logs or public sources
  • Do not promise unconfirmed publish dates
  • Do not add consulting-style CTAs
  • Do not write as if you verified something you did not verify

Success criteria

  • Passed english-tech-blog-structure and humanizer
  • Front matter or separate meta matches the shape above (published: false; at most 4 tags on the API payload)
  • For a cross-post, canonical_url points at the earlier URL
  • Includes at least one of: trade-offs, measured results, a failure story, or reproduction steps

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.