organic-growth

Organic Growth

Use this skill when an operator needs to decide what to improve on a site for organic discovery: technical search health, content demand, landing-page performance, conversion quality, or the clarity and citeability that can help a page appear in answer engines.

The output is a decision packet, not a bag of SEO tips. It separates what the evidence shows from what the analyst infers, ranks the next actions by expected impact, confidence, and effort, and says how each action should be measured. No action is executed by this skill.

Direct collection and reuse

The default review starts from an objective and the canonical site root. It collects only missing evidence, then uses the same analyze phase as a composed graph. Any supplied search, analytics or site evidence takes the reuse branch immediately: no page fetch, measurement query or extra research turn runs. Forward any prior collection_notes with those sources so access failures and measurement gaps survive the handoff. Choose analyze explicitly for a source-only graph that must never acquire data.

With no sources, the collection turn can use web-fetch#web-fetch for relevant public pages and the read-only defaults google-search-console#performance or google-analytics#report on the operator's selected compatible bindings. Google is optional; no private report is required to inspect a public page. Supply known property/date/measurement selectors or relevant page URLs in source_context. Ask when a private selector is ambiguous. A missing binding or unavailable report means unavailable evidence, not zero traffic or no commercial demand.

Collection returns actual reader packets and gaps, not a growth strategy. The existing analysis owns priorities, causal caution and measurement. For public page text, pass the unchanged provider_operation.data.result from a successful web-fetch with extract: text into site_evidence; its final URL, full body digest, extracted text and provenance survive the handoff. Preserve complete Google normalized evidence packets, including their request and coverage caveats. Never manufacture a digest, rewrite a report as live, or treat snippets as page content. Failed reads stay in collection_notes; they are not SEO observations.

A long page yields a labelled bounded text view for analysis, while its original source identity stays bound. Omitted content was not assessed. Focus another source read only if the decision needs evidence outside that view; do not turn a presentation ceiling into a claim that the site itself is unsupported.

Composes

<!-- Generated from the native execution closure; run pnpm core-skills:composes:generate. -->

• google-analytics#report
• google-search-console#performance
• web-fetch#web-fetch

Evidence comes first

The supplied-evidence analyze runner accepts a bounded objective, the site, optional business constraints, and any combination of:

• search-performance evidence, such as a sealed google-search-console performance packet;
• analytics evidence, such as a sealed google-analytics report packet;
• site or market evidence with a SHA-256 source/content digest, source reference, kind, and bounded extracted summary.

Google is optional. The canonical workflow depends on evidence contracts, not a provider, connector, credential store, or hosted tenant. A local crawler, self-hosted analytics service, another search provider, third-party connector, or Runx-hosted connector can supply evidence when its packet has a stable digest and useful provenance. Connector ids, credential handles, URLs, and tenant configuration never belong in this skill.

The admission step rejects invalid packet schemas, failed validation, missing digests, and unbounded evidence. It gives the analysis agent the rows, metadata, caveats, and business context needed for judgment, capped to an explicit analysis view. The complete source packet remains bound by digest; the model must not pretend a bounded view is exhaustive.

Every material claim and recommended action must cite one or more admitted source digests. Runx's native evidence verifier rejects invented references, changed objective/site identity, empty claims, missing decision fields, and unsupported external-effect claims. A failure becomes needs_more_evidence, not a polished guess.

Reading the combined signal

Search evidence explains discovery: queries, pages, impressions, clicks, CTR, position, freshness, and coverage. Analytics explains behavior and business outcomes under the property's measurement model. Site evidence explains what is actually published, crawlable, structured, internally linked, and supportable.

Use their disagreement. Rising impressions with weak CTR can justify inspecting snippet fit, intent, brand recognition, or SERP composition. Strong clicks with weak engagement can justify checking landing-page promise, speed, navigation, or measurement. Analytics decline without a matching Search Console decline can point toward tagging, channel classification, consent, conversion configuration, or non-Google traffic. These are hypotheses until the relevant evidence is collected.

Compare identical date windows, page identities, search types, dimensions, timezone assumptions, and definitions. Preserve incomplete Search Console data, GA4 thresholding, sampling, (other)-row loss, pagination, and imported-source status as caveats. Do not convert correlation into causation.

SEO and answer-engine work

Technical SEO recommendations require site evidence. Check indexability, canonicalization, robots rules, sitemap membership, status codes, rendering, structured data eligibility, internal discovery paths, and page duplication before prescribing a fix. Search Console URL inspection is a point observation, not authority to mutate the site or request indexing.

Content recommendations should connect a demonstrated audience need to a page, business outcome, and information gap. Avoid producing pages merely because a keyword exists. Consolidation, deletion, template repair, internal linking, or a better primary answer may beat another article.

For answer-engine or "GEO" work, optimize the underlying source quality: unambiguous entities, direct answers, original evidence, clear authorship and dates, machine-readable structure, stable URLs, strong internal context, and claims that another system can verify. There is no reliable switch for AI citations and no basis for guaranteeing inclusion. Treat answer-engine visibility as an observed channel that needs its own evidence, not as a mystical ranking score.

Action packet

A ready packet contains:

• an executive summary and bounded analysis scope;
• evidence-bound observations and explicit hypotheses;
• prioritized opportunities;
• recommended actions with rationale, source digests, owner type, effort, confidence, expected signal, stop condition, and verification method;
• a dedicated answer-engine review where evidence supports it;
• a measurement plan with baselines, leading and business outcomes, comparison windows, and rollback or reconsideration triggers;
• risks, missing evidence, and the authoritative source inventory.

The action list remains external_status: not_executed. Route implementation to the owning repository or business workflow. Content changes may move through a governed content pipeline; code and template work belongs in the site's repository workflow; publication and provider mutations keep their own approval and readback gates.

A useful operating loop

Run a narrow baseline, choose one evidence-bound action, make the change through its owner, wait for an appropriate observation window, then rerun the same queries. Preserve the original packet digests. Change several variables together only when the operator accepts that attribution will be weaker.

Stop when no source is admitted, page/property identity cannot be aligned, the date windows are incomparable, privacy or pagination caveats defeat the claim, the requested recommendation needs technical/content evidence that is absent, the agent cites a source outside the admitted set, or the objective asks for guaranteed ranking, mass low-value content, deceptive markup, or unapproved publication.

The deterministic module only admits and bounds organic-growth domain evidence. It does not hash evidence, call providers, or implement generic verification. Runx owns native digest provenance, evidence verification, agent receipts, and all downstream effect gates.

Agent task contract

organic-growth-analyze

Return growth_candidate with decision: ready, the exact objective and site_url, executive_summary, observations, hypotheses, opportunities, recommended_actions, geo_review, measurement_plan, claims, risks, missing_evidence, and external_status: not_executed.

Each claim requires a non-empty claim and admitted source_digests. Every recommended action requires non-empty source_digests plus a rationale, priority, confidence, effort, expected_signal, verification, and stop_condition. Mark statements as observed, inferred, or hypothesized. Preserve source caveats, do not invent metrics or causes, and do not claim a change, publication, provider call, ranking result, or AI citation occurred.

organic-growth-collect

Acquire relevant missing evidence for the exact objective/site, using the named read-only skill defaults and native asks if a genuine selector is missing. Do not submit sitemaps, alter analytics, publish, crawl unrelated sites or invent missing metrics. Use the returned tool schemas. Google sources are optional. Read a useful relevant public page if no source was supplied. Fetch explicit page URLs directly; avoid redundant reads and broad speculative crawling.

Return exactly search_evidence, analytics_evidence, site_evidence and collection_notes. Use empty objects/array for absent sources. Forward actual normalized Google packet data unchanged. For a text page, forward the unchanged provider_operation.data.result as an individual site-evidence record, never the provider packet wrapper. Preserve final URL, digest, text and provenance. Record inaccessible sources, actual read errors and missing measurement context as notes. Do not perform synthesis; the next analysis turn owns it. No source means honest missing evidence. Do not claim a provider call you did not make.