Seo orchestrator

Skill Hainrixz/claude-seo-ai/skills/seo-orchestrator

Orchestrates a full SEO + AI-search audit — detects the site vertical, builds a shared PageSnapshot, dispatches read-only specialist subagents in parallel, merges their findings, computes the two scores, and renders the report. Invoked by the `audit` command (the /claude-seo-ai:audit skill); not usually called directly.From its SKILL.md

Install

npx -y skills add Hainrixz/claude-seo-ai --skill seo-orchestrator

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

runs commandsInstructs the agent to run 3 commands, including `Run seo-crawl-render to build the shared PageSnapshot` and 2 more.

SKILL.md

2.9 KB, 637 tokens by cl100k_base, as published. Nobody here has run it

seo-orchestrator (Layer 2)

Coordinates the whole audit. Three phases: detect → dispatch → synthesize.

1. Detect

Resolve the target (URL or local path).
Run seo-crawl-render to build the shared PageSnapshot (raw HTML, rendered DOM when available, response headers, status chain, plus site artifacts: robots.txt, sitemap set, llms.txt). Record the data tier reached (0/1/2).
Run seo-vertical-detect to classify the target. Read references/routing.md to get the always-on + conditional module set and the subagent dispatch table.

2. Dispatch (parallel)

Spawn the read-only specialists in parallel (one message, multiple Task calls) so their verbose intermediate output stays isolated. Pass each the PageSnapshot location/contents, the detected vertical, and its assigned modules:

technical-auditor → M1, M2 (+M3), M4, M7, M7b (mobile), M7c (headings), M8, M9, M10, M15, M17
ai-search-geo-specialist → M6, M11, M12, M14, M21
content-eeat-analyst → M13, M16
schema-generator → M5 (and M18/M19 schema when those verticals are active)

Each subagent returns an array of findings conforming to schema/finding.schema.json. Subagents are read-only (no Write/Edit tool) — the audit can never mutate files.

3. Synthesize

Merge all findings; dedupe by id (keep the most severe status).
Run seo-score over the merged findings to get the two scores with category breakdowns and severity gating.
Render the report (see references/scoring-model.md for bands/interpretations):
- Header: Search SEO band/score + AI Visibility band/score, each with a one-line interpretation.
- Per-category table for each score (value, weight, active?).
- Data confidence: tier reached + count of needs_api findings.
- Prioritized actions: top findings sorted by severity × magnitude ÷ effort, each showing status, evidence, recommendation, fixability (auto/proposed/advisory), and expected_impact (axis + confidence + magnitude).
- Footer: how to run /claude-seo-ai:fix.

Notes

Keep this skill's own output tight — the value is the merged report, not narration.
If the target is a multi-page crawl, analyze the homepage + one instance of each detected template first, then sample additional pages; log what was sampled (never silently cap coverage).
Degrade gracefully: if a render/PSI/GSC tier is unavailable, proceed at the lower tier and mark affected findings needs_api.

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Gives 0 of the 12 instructions most agent orchestration skills give in 637 tokens

Counted across 848 of the 1,300 authors here whose files we hold, read 2026-09-06

Dispatch one agent per independent problem domainin 56 of 848, across 42 files
Run full test suite after integrationin 55 of 848, across 42 files
Verify fixes do not conflictin 40 of 848, across 32 files
Review each summary when agents returnin 40 of 848, across 31 files
Write a handoff document summarising the current conversationin 30 of 848, across 25 files
Reference existing artifacts by path or URLin 26 of 848, across 24 files
Give each agent a specific scopein 19 of 848, across 10 files
Give each agent a clear goalin 19 of 848, across 10 files
Include a suggested skills section in the documentin 18 of 848, across 16 files
Tailor the doc to the user argumentsin 18 of 848, across 15 files
Issue all subagent dispatches in the same responsein 17 of 848, across 11 files
Use git worktrees for isolationin 17 of 848, across 8 files

Said here and by no other author read

resolve the target
render the report
Run seo-crawl-render to build PageSnapshot
Run seo-vertical-detect to classify the target
Spawn the read-only specialists in parallel
Merge all findings and dedupe by id

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.