dineway/dineway-skills

dineway-seo

Plan and execute Dineway SEO, local SEO, GEO, and answer-engine visibility workflows. Use when creating content briefs, topic clusters, competitor/comparison pages, schema JSON-LD, SEO image assets, or Firecrawl/DataForSEO research through Dineway Tools.

First seen Jun 16, 2026

Installation

$ npx skills add dineway/dineway-skills --skill dineway-seo

Also in this package

Other skills from dineway/dineway-skills · top by installs.

npx skills add dineway/dineway-skills

Browse all from dineway/dineway-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 Not 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 2
License LICENSE
Default branch main
Open issues 0
Status Active

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 5,475 B
  • docs SUMMARY.md 273 B

History

  1. First seen on skills.sh
  2. First recorded snapshot · 11 installs

SKILL.md

Dineway SEO

Dineway SEO is a skill-first workflow for search, local, and answer-engine optimization. Use it to gather evidence, design changes, and apply them through existing Dineway content, media, SEO metadata, schema/settings, rendering, and code-patch paths.

Boundaries

  • Do not require third-party Agent plugins, slash commands, subagents, or provider MCP servers. Use

Dineway CLI/MCP and the current Agent's native capabilities. Never ask for secret values; Dineway Tools may use explicit CLI environment references or encrypted site BYOK.

  • Crawl the first-party site locally and use Browser Use when rendering is required. Keep any

deterministic crawler control bounded and separate from Agent conclusions.

  • Route live provider data only through the Dineway Tools Discover → Inspect → Run flow. Prefer

dineway tools ...; use the generic tools_* MCP operations when CLI is unavailable. Require tools:read and tools:run for token-authenticated MCP access.

  • Use raw Firecrawl/DataForSEO payloads returned by Forgeway, but do not expose provider credentials or billing secrets.
  • Image generation is a skill workflow. Use the current agent image capability only when the user explicitly asks for generated images; do not consume Forgeway AI quota for SEO images.
  • Apply generated work as a Draft. Publish or schedule only after exact-Draft editorial approval,

valid release authorization, and current policy/capacity checks.

Child Skills

Use this root skill for broad, mixed, or ambiguous SEO/GEO work. When the task clearly matches one workflow, use the matching child skill directly:

  • dineway-content-brief: content briefs, outlines, page refresh plans, metadata recommendations.
  • dineway-seo-cluster: topic clusters, hub/spoke plans, internal link matrices.
  • dineway-seo-competitor-pages: alternatives, comparison, roundup, and feature/price comparison pages.
  • dineway-seo-schema: schema/JSON-LD audits, recommendations, validation, and Dineway schema apply plans.
  • dineway-seo-images: SEO image planning, generation prompts, alt text, filenames, media apply plans.
  • dineway-seo-providers: Firecrawl/DataForSEO research through Dineway Tools, provider choice, and evidence interpretation.

Child skills must still use the shared Dineway apply loop and provider/security boundaries in references/apply-loop.md, references/security-boundaries.md, references/provider-selection.md, references/firecrawl.md, and references/dataforseo.md when relevant.

Workflow

  1. Identify the job type and load only the relevant reference:

- content brief: references/content-brief.md - topic cluster: references/topic-cluster.md - competitor/comparison page: references/competitor-pages.md - FLOW operating model: references/flow-framework.md - schema/JSON-LD: references/schema-jsonld.md - image asset planning: references/image-generation.md - Firecrawl research: references/firecrawl.md - DataForSEO research: references/dataforseo.md - provider choice: references/provider-selection.md - provider/security boundaries: references/security-boundaries.md - applying changes: references/apply-loop.md

  1. Inventory current Dineway state before proposing changes:

- collections and hasSeo support; - current content item, slug, title, excerpt, SEO metadata, status, and revision; - site SEO settings, canonical/base URL, locale/hreflang, and JSON-LD contributors; - relevant media assets, alt text, captions, and public URLs; - site-context briefing, brand voice, policies, seasonal strategy, and human-in-the-loop notes.

  1. Decide whether local/Browse evidence is sufficient. Use Dineway Tools discovery before selecting

any endpoint. Use DataForSEO for managed SERP, keyword, competitor, maps, and AI visibility datasets; use Firecrawl for infrastructure-scale page extraction, URL discovery, crawl jobs, and search+scrape.

  1. Produce a finding set before changes:

- source evidence and dates; - prioritized opportunities; - skipped/risky claims; - Dineway write targets; - verification steps.

  1. Apply through the narrowest existing surface:

- content CRUD for page/post/menu/review/news/gallery copy; - SEO metadata fields for title, description, OG, canonical, noindex, and robots; - media upload/update for images, alt, captions, and dimensions; - site settings and SEO Graph contributions for global metadata and schema; - Astro code patches only when Dineway content/settings cannot express the needed rendering.

  1. Verify rendered output and data consistency. Separate applied changes from recommendations that still need human approval or unavailable data.

Output Contract

For planning tasks, return:

  • Goal
  • Evidence Used
  • Search Intent / Audience
  • Recommended Dineway Changes
  • Provider Calls Used
  • Skipped or Risky Claims
  • Verification Plan

For applied tasks, return:

  • Applied Changes
  • Files or Content Items Changed
  • Provider Calls Used
  • Verification Results
  • Remaining Risks

Do not include internal implementation instructions in public-facing copy.