| name | technical-seo-audit |
| description | Audit production SEO — metadata, canonical/OG, sitemap, robots, structured data, and Core Web Vitals signals. Use when auditing a live site or filing SEO recommendations. Creates GitHub issues labelled `type:seo-recommendation` with owning squad. Do NOT implement fixes — hand to engineering squads.
|
| license | MIT |
| allowed-tools | Read Write Glob Grep Bash |
| argument-hint | [--url base-url] [--focus metadata|sitemap|structured-data|cwv] |
| metadata | {"author":"Carinya Parc","version":"0.1.0","owner":"search-optimisation","review_cadence":"quarterly","work_shape":"monitor-and-report","output_class":"tracking-update"} |
Technical SEO audit
You audit the production site and file prioritised recommendations as GitHub
issues. Pass optional --url and --focus after the skill name. Do not open
PRs or edit application code.
Read search-optimisation-conventions.md.
Inputs
| Input | Location | Required |
|---|
| Audit URL | Target production URL or --url | Yes |
| Focus area | --focus or full audit | No |
| robots/sitemap | Public paths + repo config | If present |
| Squad labels | Instance squad charter | If present |
Steps
- Resolve target via
.agency/target.json or instance config. Default URL from
target production URL (or --url).
- Read public
robots.txt and sitemap config in repo when available. Read squad
charter for label conventions when present.
- Audit via Playwright (live) and repo files per the checklist below.
- Create GitHub issues via the github connector, priority-ordered.
- If no issues were created, write an audit trail to
.agency/work/seo/technical-seo-audit-{YYYY-MM-DD}.md.
Checklist
Metadata (sample key pages)
Homepage, primary index pages, and sample detail pages per content type:
<title> unique and under ~60 chars
<meta name="description"> present, unique, under ~160 chars
<link rel="canonical"> correct
- Open Graph:
og:title, og:description, og:image, og:url
Sitemap and robots
robots.txt allows indexing of public content
- Sitemap reachable and lists published content
- No stale or 404 URLs in sitemap
Structured data
- Content-type pages use appropriate schema (Article, Product, Recipe, etc.)
- Note rich-results errors in issues
Core Web Vitals (lab)
- Note LCP/CLS/INP concerns from Playwright performance metrics when available
- Flag render-blocking or missing image dimensions as recommendations
Issue format
| Priority | Examples |
|---|
| P1 | Indexing blockers, missing canonicals, broken sitemap |
| P2 | Missing OG tags, incomplete structured data |
| P3 | CWV improvements, metadata polish |
Each issue:
- Title:
[SEO] {short description}
- Labels:
type:seo-recommendation, owning squad label per conventions
- Body: evidence, impact, recommended fix, page URL
Quality rules
- Each issue has evidence (URL, screenshot note, or HTML snippet)
- Owning squad label assigned per conventions
- Actionable fix description for engineering — implementation is out of scope