Workflow ยท one run, two views

Reuse the fanout. Do not run it twice.

AI Fanout owns the provider request. SEO Fanout receives a bounded copy of the result already visible to the visitor and performs only deterministic browser-side analysis.

AI Fanout

Observe the search actions.

The upstream tool records the short input, provider, locale, exposed query strings and sources at the scope returned by the API. It does not claim consumer-interface internals.

SEO Fanout

Inspect the research shape.

The downstream view groups query wording, counts recurring terms, preserves source scope and identifies checks still required before treating the run as SEO evidence.

Data flow

The handoff remains in the browser.

  1. 01

    Select the useful visible queries

    AI Fanout starts with all exposed queries selected. The visitor may narrow the set before continuing.

  2. 02

    Create a versioned fragment

    The browser creates ai-fanout.seo-research-handoff/1.0. Raw provider responses, secrets and local history are excluded.

  3. 03

    Validate and clear the fragment

    SEO Fanout validates size, schema, producer, URLs and field limits, then removes the fragment from the address bar.

  4. 04

    Build the local plan

    No API call occurs. The user may copy or download the resulting research plan.

Evidence boundary

Observed query strings are still only one bounded run.

They do not establish demand, ranking factors, permanent provider behaviour or a requirement to create content. Source relationships remain exactly as coarse as the provider exposed them.

Start one AI Fanout run