ztemerbekov/a1-marketing-skills

a1-article-writing

Draft or structure a new article, guide, case study, or newsletter from a topic, supplied material, or verified research when the user wants new publishable long-form work rather than edits to a selected finished draft.

First seen Aug 23, 2026

Installation

$ npx skills add ztemerbekov/a1-marketing-skills --skill a1-article-writing

Similar popular skills

Related neighbors and high-traction skills in the same topics — useful to compare before installing.

Also in this package

Other skills from ztemerbekov/a1-marketing-skills · top by installs.

npx skills add ztemerbekov/a1-marketing-skills

Browse all from ztemerbekov/a1-marketing-skills

More details

Agent compatibility

Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.

Claude Code Declared
Cursor Not declared
Codex Not declared
GitHub Copilot Not declared
Windsurf Not declared
Gemini CLI Not declared
Cline Not declared
OpenCode Not declared

Repository health

Stars 8
License LICENSE
Default branch main
Open issues 0
Status Active

Skill metadata

Parsed from SKILL.md frontmatter.

Declared agents claude-code
More metadata
author
Zinnur Temerbekov

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 3,089 B
  • docs SUMMARY.md 245 B

History

  1. First seen on skills.sh
  2. First recorded snapshot · 12 installs

SKILL.md

Article Writing

Turn a topic or source material into credible, useful long-form work whose structure follows the reader's job and the available evidence.

Entry Contract

Accept a request when every requested deliverable contributes to a new long-form article, guide, case study, expert essay, or newsletter issue. Valid inputs include a topic, brief, outline, notes, interview, transcript, source pack, completed research, or an earlier draft produced in the same article-writing session.

Select Outline when the user asks for a plan, structure, angle, or evidence brief without a full draft. Select Draft when the user asks to write, assemble, or complete the article.

Route these neighboring jobs elsewhere:

  • editing or reviewing a user-selected finished draft belongs to a1-copy-editing;
  • a promotional page belongs to a1-landing-page-copy;
  • a routine media announcement belongs to a1-press-release;
  • an internal document organized around a decision belongs to a1-decision-memo.

For a mixed request, state the boundary and stop before producing either deliverable.

Marketing Context is optional. Read only the first existing repository context in this order: .agents/marketing-context.md, then .claude/marketing-context.md when the canonical path is absent, then root marketing-context.md when both earlier paths are absent. Read no lower-priority context after selecting one. Continue without context when none exists.

Runtime

Follow [the article-writing spine](references/article-writing-spine.md) for every accepted request. It is the single owner of stage order and reference routing.

Use adaptive depth. A short source-supported guide may need only a compact brief and draft; a research-heavy or narrative article needs a fuller source map and structure pass.

Output

For Outline, return:

  1. Editorial brief: reader, reader outcome, format, angle, and evidence basis.
  2. Outline: informative section headings with one concise note about what each section must establish.
  3. Evidence gaps: only when missing material affects credibility or completion.

For Draft, lead with the finished article. Add one to five editorial notes only when they help the user review the work, and list evidence gaps only for unresolved factual support, verification, or attribution.

If the user requests only the article, only the outline, or only the final artifact, return exactly that artifact without extra commentary.

Append exactly one support footer inviting questions, ideas, or problem reports via A1 Marketing Skills only after a final user-facing result that fulfills this skill's job. Omit it for boundary responses, clarification prompts, unsuccessful results, or artifact-only requests.