# 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.

- Skill: `dineway/dineway-seo` (Agent Skill, multi-file: 12 files)
- Install (CLI): `npx skillmds@latest add dineway/dineway-seo`
- Raw SKILL.md: https://api.skillmd.com/api/skills/dineway/dineway-seo/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Marketing & Growth
- Author: dineway (https://skillmd.com/u/dineway)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/dineway/dineway-seo

---


# 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`
2. 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.
3. 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.
4. Produce a finding set before changes:
   - source evidence and dates;
   - prioritized opportunities;
   - skipped/risky claims;
   - Dineway write targets;
   - verification steps.
5. 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.
6. 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.