agentleFS
XPOZpublicXPOZpublic/geo-seo-agent/.claude/skills/content-generation/SKILL.md

content-generation

Skill0 starsChanged 3 months agoMITView on GitHub

Turn one approved brief into a publish-ready content draft. Fresh verified research, GEO-formatted markdown in content/self-published/items/<slug>/, handed to the human (or the company's publishing workflow) to publish. Never chooses topics; never publishes.

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

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

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

    Use the session's web search tools (WebSearch, WebFetch) for all research, never the ask scripts.

    line 27

≈ “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

Content Generation

Turns exactly one approved brief into one publish-ready draft. The strategy (what to write and why) comes entirely from the brief; this skill supplies substance (fresh verified research) and craft (GEO formatting). Because every site publishes differently, the deliverable is a complete, CMS-ready markdown draft plus everything the publisher needs; the actual publish step follows the workflow recorded in strategy/company-profile.md, or stays with the human when none is defined.

Boundaries

  • Requires an approved row with action: new-content and a brief.md in the item folder its channel defines: content/self-published/items/<slug>/ (row in content/index.md) or content/docs/items/<slug>/ (row in content/docs/index.md). No approved brief → stop; do not invent topics.
  • One brief per run. Never publishes or deploys; publishing is always the human's act (directly or by merging a PR the workflow defines).

Process

1. Read the brief

Target prompt IDs, content type, angle, key points, primary + secondary keywords, meta description. Validate status is approved.

2. Research fresh

Before searching, check memory/competitors/ for entity files on the tools and providers relevant to this brief. Entity files contain source-verified pricing and positioning with URLs; use them as starting context. Still verify facts by web search at write time; entity files carry a verification date per source, and anything older than ~4 weeks warrants re-verification.

Every factual claim (competitor pricing, features, statistics, dates) is verified by web search at write time, with sources noted. Nothing recalled from memory goes into the draft. Comparison content states facts about other companies; a wrong claim is a liability and a GEO credibility killer. Your own product facts (pricing, capabilities, positioning) come from strategy/company-profile.md, stated the way that file frames them.

Use the session's web search tools (WebSearch, WebFetch) for all research, never the ask scripts. Those exist for visibility measurement, and they are significantly more expensive per call.

3. Draft

Write draft.md into the item's folder: a metadata block the publisher can map onto their CMS or docs system, then the full body. For docs items, follow the docs system's own page format (recorded in the company profile's publishing workflow) and keep the metadata block to what that system uses.

---
title: <working title>
seoTitle: <50-60 chars, primary keyword first>
description: <150-160 chars, action-verb start>
keywords: [<6-8, at least 2 long-tail, no generic single words>]
date: '<YYYY-MM-DD today>'
dateModified: '<YYYY-MM-DD today>'
targetPrompts: [<panel IDs>]
---

Body: follow strategy/geo-writing-principles.md for opening structure, brand placement, extractability, and tone. Self-contained paragraphs (40-75 words), comparison tables with concrete data, question-shaped FAQ headings ("What is...?", "How to...?"), statistics with named sources.

4. Hand off

If the company profile defines a publishing workflow (a CMS to paste into, a site repo to open a PR against), follow it exactly, stopping at the human gate it defines. Otherwise, present the draft and a short publish checklist: where it should live (URL slug), that the page should carry Article/FAQ structured data if the site supports it, and that dateModified must be kept honest on future edits.

5. Record

Flip the item's row in its channel index to in-production (or published with the URL, once the human confirms it's live), append the journal entry, refresh memory/state.md if state changed.

Known limitations

Single-pass drafting (no separate editorial reviewer). If the site's publishing workflow uses hero images, generate one with /generate-hero-image after the draft is final and include it in the handoff.

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.