agentleFS
eunomia-bpfeunomia-bpf/eunomia.dev/.agents/skills/medium-publisher/SKILL.md

medium-publisher

Skill244 starsChanged 5 days agoMITView on GitHub

Prepare or publish eunomia.dev Markdown articles on Medium through faithful long-form syndication. Use when asked to paste, import, publish, QA, or record Medium posts for blog, tutorial, project, paper, release, or GitHub content. Preserves the English source title and body exactly apart from mechanical rendering fixes, with optional canonical settings, publication authorization, browser QA, follow-up, and media ledger updates.

  • Reads credentials

The file

Show its rules and text

We split this skill into the rules it gives an agent

13 rules. About 0 are common, 8 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

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

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

    If the source path is missing, inspect .github/publisher/posts_queue.txt, .github/publisher/media/not-published.md, and recent …

    line 19
  2. Distinctive:

    Import story: <https://medium.com/p/import>

    line 27
  3. Distinctive:

    Use a browser surface with the logged-in session when UI work is required.

    line 31
  4. Distinctive:

    If a creation request times out or returns an unknown result, check the public profile for the exact title before retrying.

    line 54
  5. Distinctive:

    Read the canonical English source and record the source URL, GitHub links, and paper links when known.

    line 61
  6. Distinctive:

    If the source is not suitable for Medium, skip it or fix the source first.

    line 71
  7. Very specific:

    If Medium rejects otherwise valid Markdown with parser error 2012, convert the prepared artifact to semantic HTML and publish with …

    line 76
  8. Very specific:

    Before sending the Medium API request, write or update the Medium draft record under draft/media/YYYY-MM-DD/<source-slug>/medium.md using …

    line 83
  9. Distinctive:

    no confidential or unreleased claims appear

    line 103
  10. Distinctive:

    Before confirmed publishing, inspect the complete local upload artifact.

    line 106
  11. Distinctive:

    After confirmed publish and public-page QA, capture the Medium URL.

    line 127
  12. Very specific:

    After confirmed publish, update .github/publisher/media/published.md with source path, canonical URL, Medium URL, date, tags/publication, …

    line 134
  13. Very specific:

    Before final completion, add any Medium-specific issue encountered during this session to this skill or references/platform-preferences.md, …

    line 138

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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

Medium Publisher

Prepare and publish Medium stories from canonical eunomia.dev content through the Medium API. Publication authorization follows CLAUDE.md's Precedence Rule and Publishing section.

Inputs

  • Source Markdown path or source/canonical eunomia.dev URL.
  • Source title, optional source subtitle, publication, tags, cover, or optional canonical URL.
  • Optional GitHub link, paper link, or follow-up plan.

If the source path is missing, inspect .github/publisher/posts_queue.txt, .github/publisher/media/not-published.md, and recent docs/blog/posts/.

Platform Entry Points

Use a browser surface with the logged-in session when UI work is required.

References

Load references/platform-preferences.md when choosing canonical syndication settings, light Medium metadata/rendering adaptations, title/subtitle/cover, tags, browser QA, promotion balance, or follow-up. Do not load broad strategy drafts for routine publishing unless the user asks for campaign or content-platform planning.

API-First Platform Boundary

Publish through the documented Medium API by default, using MEDIUM_API_KEY from the local .env or an approved secret-backed publisher. Never print, record, or commit the token. Verify the authenticated account with /v1/me, check the public profile for an exact-title duplicate, and create the story with POST /v1/users/{authorId}/posts.

The Medium API is archived and unsupported, so treat every response as untrusted until the public story has been checked. Do not use private or hidden Medium endpoints. Use the visible web editor only for repairs the API cannot perform, then repeat public-page QA.

If a creation request times out or returns an unknown result, check the public profile for the exact title before retrying. Make at most one retry; if its result is also unknown, stop, leave the queue item unfinished, and record the exception instead of risking duplicate stories.

Draft Preparation

  1. Read the canonical English source and record the source URL, GitHub links, and paper links when known.
  2. Preserve the source title exactly and keep the body substantively unchanged.
  3. Make only mechanical adaptations: remove site front matter or a duplicate H1, convert heading levels, repair image URLs/uploads, and preserve readable code, tables, formulas, embeds, and links. Set Medium tags, publication, cover, and other metadata without changing the article.
  4. Preserve GitHub, project, paper, and source links already in the article. Add body text or links only after the same source change or an explicit user request.
  5. If the source is not suitable for Medium, skip it or fix the source first. Rewrite, translate, shorten, expand, reorder, or split it only when the user explicitly asks for that specific publication.
  6. Include the visible article H1 inside the API content; the API title field controls listing and SEO metadata but does not render the story title. If Medium rejects otherwise valid Markdown with parser error 2012, convert the prepared artifact to semantic HTML and publish with contentFormat: "html". Keep no more than three tags because Medium ignores additional tags.

Draft Archive

Before sending the Medium API request, write or update the Medium draft record under draft/media/YYYY-MM-DD/<source-slug>/medium.md using the local date. For canonical imports, this file may reference the source Markdown body instead of duplicating it, but it must record the exact source title, optional source subtitle, canonical relationship when configured, GitHub/paper links, tags, source/project note if useful, media choices, and QA state. For long-form posts, finish the Medium-specific artifact locally before publishing. Use the web editor only for supported metadata changes or repairs that remain necessary after creation.

Browser QA

  • canonical/import relationship is correct when configured
  • the Medium body has not drifted from the canonical article except for necessary formatting/link/tag edits
  • title matches the source exactly; subtitle is unchanged from the source or omitted unless the user requested one
  • code blocks, images, embeds, links, and headings render cleanly
  • tags are relevant and not spammy
  • no confidential or unreleased claims appear
  • the API request has not been sent

Before confirmed publishing, inspect the complete local upload artifact. After confirmed publishing, open the public Medium URL and scroll through the rendered story from top to bottom before updating the ledger. Verify images actually load, title/subtitle are not polluted by site suffixes, headings do not include empty artifacts, tables have survived or have readable fallbacks, code blocks are not mangled by language detection labels, canonical settings when configured, source/project links work, and mobile/narrow rendering is usable when practical. If the public page exposes a formatting issue, edit the published story through the web UI and repeat the public-page check.

Medium API conversion or import is allowed to preserve the canonical body, but it is not safe to trust blindly. Specifically check whether titles carried the source site suffix such as | eunomia, whether Markdown tables were flattened into loose paragraphs, whether image captions are empty placeholders, and whether code block language labels appeared as prose. If Medium cannot preserve a table cleanly, replace that table with a readable list or compact prose fallback in the web editor before publishing.

Follow-Up

After confirmed publish and public-page QA, capture the Medium URL. Monitor responses, highlights, private notes, and publication feedback only when the user asks or follow-up was part of the task. Draft replies that add context rather than sell.

Ledger Update

After confirmed publish, update .github/publisher/media/published.md with source path, canonical URL, Medium URL, date, tags/publication, media, and follow-up notes.

Before final completion, add any Medium-specific issue encountered during this session to this skill or references/platform-preferences.md, then record the public-page QA result in the draft record. In particular, check that API-created stories visibly contain their H1 and that HTML conversion preserved images, headings, code blocks, links, and readable table fallbacks.

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.