SEO Scale Agent

⬜ Proposed — not built. Tier 2 — one dependency

What it would do

Works out which market x topic combinations deserve a page, down to micromarket level, and outputs ranked page briefs rather than mass-generating pages.

Why an agent

The combinatorics are mechanical. Judging which combinations justify a page — and which would be thin duplicates — is not.

What it needs first

A page-generation pipeline, and ideally Search Console. Discovery already tags every item with an Indian state, so the input half exists.

Roadmap source

Scale Growth — Discovery to Nat / State / City / Micromarket (page 3) — see roadmap_ref_extracted.md.

Design notes and open questions

Thin-content risk is the headline caution. Mass-generating micromarket pages is a known search-penalty vector. The agent must flag cannibalisation and refuse weak combinations, not maximise page count.
Outputs briefs for human approval rather than publishing directly.
Needs the published-page inventory to detect overlap; that inventory does not exist yet.

Subagents

Market ranker — which market x topic combinations justify a page at all
Cannibalisation checker — does this compete with a page we already publish
Brief writer — outputs a brief for approval, not a published page
Thin-content refuser — the veto that stops the agent maximising page count

Stub spec. Expand to the full overview.md + technical.md split (as the three built agents use) when this moves from proposed to scheduled.