​

sources

orc sources は、Workspaceの回答からsource domain・URL・引用の集計と、競合を含む回答のsource gapを調べるコマンドです。domain、host、URL単位の表示を使い分け、期間・brand・topic・tag・platformで対象の回答を絞れます。

保存済みの取得根拠と引用を対象にし、新しい観測を開始しません。--project は互換queryですが、現在のendpointはproject filterを適用しません。対象を絞るにはbrand、topic、prompt tagなどの対応filterを使います。

source domainを一覧表示します。

開始と終了は含むUTC日付 YYYY-MM-DD です。top は現在の取得または引用があるsource、new は期間内に初取得されたsource、trending と losing は取得数の増減です。cohortは重複可能です。trending/losingは両方の日付が必要で、直前の同じ日数の期間と比較します。両方の境界がなければ空になります。

domain一覧のsort既定はcohortなしでcitation_count、topでretrieval_count、newでfirst_seen_at、trending/losingでretrieval_changeです。losingは既定昇順、他は降順で、--order で変更できます。

回答に含まれる異なる設定済み競合の数を --min-competitors で指定できます。--view はdomain・host・urlで、既定domainです。指定した言及brandの条件は --mentioned-brand-operator or / and でany/allを選びます。hostname・URL・titleのsearchは大文字小文字を区別しません。

Grouping axis: domain combines subdomains under their registrable domain; host keeps captured hostnames; url keeps individual citation URLs. Defaults to domain.; enum: domain|host|url

型: string。任意。

Minimum distinct configured competitors found in an answer.

型: number。任意。

Inclusive UTC start date (YYYY-MM-DD). Omit for no lower bound.

型: string。任意。

Inclusive UTC end date (YYYY-MM-DD). Omit for no upper bound.

型: string。任意。

Case-insensitive source hostname, URL, or title search.

型: string。任意。

Comma-separated platform identifiers, such as openai or anthropic. Values are trimmed and lowercased. Matches any selected platform; omit or send an empty value for no platform filter. Combines with other filter fields using AND.

型: string。任意。

Comma-separated topic IDs for the answer population. Matches any selected topic in the workspace. Omit or send an empty value for no topic filter. Combines with other filter fields using AND.

型: string。任意。

Comma-separated IDs of the brands that own the answer prompts. Matches any selected parent brand in the workspace; this does not select every brand mentioned in an answer. Omit for no parent-brand filter. Combines with other filter fields using AND.

型: string。任意。

Comma-separated prompt tag names, not tag IDs. Values are trimmed and lowercased. Matches any selected name. Omit or send an empty value for no tag filter. Combines with other filter fields using AND.

型: string。任意。

Comma-separated competitor Brand IDs that must be mentioned in qualifying answers.

型: string。任意。

Comma-separated ISO 3166-1 alpha-2 country codes for qualifying prompts.

型: string。任意。

Whether any or all selected mentioned Brand IDs must occur in a qualifying answer.; enum: or|and

型: string。任意。

Comma-separated SourceDomainClassification values.

型: string。任意。

Comma-separated SourceUrlClassification values.

型: string。任意。

Field used to order the derived gap rows.; enum: source|competitor_count|retrieval_count|retrieved_percentage|retrieval_rate|citation_rate|gap_score

型: string。任意。

Sort direction.; enum: asc|desc

型: string。任意。

One-based page number. Defaults to 1; values above 100000 are capped. Reuse the same filters when incrementing the page.

型: number。任意。

domain viewはPublic Suffix Listでsubdomainを登録可能domainに集約し、host viewは取得した正規化hostnameを維持します。domainとhostでclassificationの語彙が異なります。--page は1始まりで既定1、100000より上はcapします。

Compatibility query parameter; this endpoint does not apply a project filter. Use the supported brand, topic or prompt-tag filters to narrow the workspace data.

型: string。任意。

Grouping axis: domain combines subdomains under their registrable domain using the Public Suffix List; host keeps normalized captured hostnames. Defaults to domain.; enum: domain|host

型: string。任意。

Inclusive UTC start date (YYYY-MM-DD). Omit for no lower bound. For trending or losing, provide both start_date and end_date. The comparison period immediately precedes the selected period and has the same inclusive day count. Without both bounds these cohorts return no rows.

型: string。任意。

Inclusive UTC end date (YYYY-MM-DD). Omit for no upper bound. For trending or losing, provide both start_date and end_date. The comparison period immediately precedes the selected period and has the same inclusive day count. Without both bounds these cohorts return no rows.

型: string。任意。

Optional source subset: top includes any current retrieval or citation; new requires a retrieval and first retrieval within the selected period; trending and losing require a positive or negative retrieval change. Cohorts can overlap. For trending or losing, provide both start_date and end_date. The comparison period immediately precedes the selected period and has the same inclusive day count. Without both bounds these cohorts return no rows.; enum: top|new|trending|losing

型: string。任意。

Comma-separated platform identifiers, such as openai or anthropic. Values are trimmed and lowercased. Matches any selected platform; omit or send an empty value for no platform filter. Combines with other filter fields using AND.

型: string。任意。

Comma-separated topic IDs for the answer population. Matches any selected topic in the workspace. Omit or send an empty value for no topic filter. Combines with other filter fields using AND.

型: string。任意。

Comma-separated IDs of the brands that own the answer prompts. Matches any selected parent brand in the workspace; this does not select every brand mentioned in an answer. Omit for no parent-brand filter. Combines with other filter fields using AND.

型: string。任意。

Comma-separated prompt tag names, not tag IDs. Values are trimmed and lowercased. Matches any selected name. Omit or send an empty value for no tag filter. Combines with other filter fields using AND.

型: string。任意。

Comma-separated classification values. Matches any supplied value; omit to include all classifications. In host view use host-role values; in domain view use domain classification values.

型: string。任意。

Sort field. When omitted: citation_count without cohort; retrieval_count for top; first_seen_at for new; retrieval_change for trending or losing. retrieval_change requires cohort.; enum: hostname|classification|retrieval_count|retrieval_change|retrieval_rate|retrieval_frequency|citation_count|citation_rate|url_count|share_of_voice|first_seen_at|last_seen_at

型: string。任意。

Sort direction. When omitted, losing uses ascending order and all other cases use descending order. Set explicitly to override the cohort default.; enum: asc|desc

型: string。任意。

One-based page number. Defaults to 1; values above 100000 are capped. Reuse the same filters when incrementing the page.

型: number。任意。

host viewの一覧で返されたsource-domain hostnameを指定して詳細を取得します。domain集約の表示値と区別してください。

source URLを一覧表示します。cursorを返されたまま使い、期間とfilterを維持します。cohortを指定しない既定sortはcitation_countです。retrieval_count・retrieval_changeによるsortにはcohortが必要です。

Compatibility query parameter; this endpoint does not apply a project filter. Use the supported brand, topic or prompt-tag filters to narrow the workspace data.

型: string。任意。

Inclusive UTC start date (YYYY-MM-DD). Omit for no lower bound. For trending or losing, provide both start_date and end_date. The comparison period immediately precedes the selected period and has the same inclusive day count. Without both bounds these cohorts return no rows.

型: string。任意。

Inclusive UTC end date (YYYY-MM-DD). Omit for no upper bound. For trending or losing, provide both start_date and end_date. The comparison period immediately precedes the selected period and has the same inclusive day count. Without both bounds these cohorts return no rows.

型: string。任意。

Optional source subset: top includes any current retrieval or citation; new requires a retrieval and first retrieval within the selected period; trending and losing require a positive or negative retrieval change. Cohorts can overlap. For trending or losing, provide both start_date and end_date. The comparison period immediately precedes the selected period and has the same inclusive day count. Without both bounds these cohorts return no rows.; enum: top|new|trending|losing

型: string。任意。

Comma-separated platform identifiers, such as openai or anthropic. Values are trimmed and lowercased. Matches any selected platform; omit or send an empty value for no platform filter. Combines with other filter fields using AND.

型: string。任意。

Comma-separated topic IDs for the answer population. Matches any selected topic in the workspace. Omit or send an empty value for no topic filter. Combines with other filter fields using AND.

型: string。任意。

Comma-separated IDs of the brands that own the answer prompts. Matches any selected parent brand in the workspace; this does not select every brand mentioned in an answer. Omit for no parent-brand filter. Combines with other filter fields using AND.

型: string。任意。

Comma-separated prompt tag names, not tag IDs. Values are trimmed and lowercased. Matches any selected name. Omit or send an empty value for no tag filter. Combines with other filter fields using AND.

型: string。任意。

Comma-separated classification values. Matches any supplied value; omit to include all classifications. For citation aggregates, classification_scope selects the vocabulary.

型: string。任意。

Sort field. When omitted: citation_count without cohort; retrieval_count for top; first_seen_at for new; retrieval_change for trending or losing. retrieval_count and retrieval_change require cohort.; enum: url|hostname|retrieval_count|retrieval_change|citation_count|share_of_voice|first_seen_at|last_seen_at

型: string。任意。

Sort direction. When omitted, losing uses ascending order and all other cases use descending order. Set explicitly to override the cohort default.; enum: asc|desc

型: string。任意。

一覧で返されたURLをURL encodeした値で詳細を取得します。新しいsourceの登録やURLのcrawlを要求する操作ではありません。

--group-by にdomain、hostname、url、classification、dateをカンマで指定します。既定のgroupはありません。dateを含む場合は --bucket-width day / week / month が必要です。classification scopeはdomain・host・urlを選択します。

Compatibility query parameter; this endpoint does not apply a project filter. Use the supported brand, topic or prompt-tag filters to narrow the workspace data.

型: string。任意。

Inclusive UTC start date (YYYY-MM-DD). Omit for no lower bound.

型: string。任意。

Inclusive UTC end date (YYYY-MM-DD). Omit for no upper bound.

型: string。任意。

One or more grouping axes separated by commas: domain, hostname, url, classification or date. No default. Use bucket_width when date is included.

型: string。必須。

Calendar date bucket width. Required when group_by includes date; otherwise omit it.; enum: day|week|month

型: string。任意。

Selects whether group_by=classification returns domain classifications or URL classifications.; enum: domain|host|url

型: string。任意。

Comma-separated hostnames.

型: string。任意。

Comma-separated registrable domains; subdomains are normalized to their registrable domain.

型: string。任意。

Comma-separated URLs.

型: string。任意。

Comma-separated classification values. Matches any supplied value; omit to include all classifications. For citation aggregates, classification_scope selects the vocabulary.

型: string。任意。

Comma-separated platform identifiers, such as openai or anthropic. Values are trimmed and lowercased. Matches any selected platform; omit or send an empty value for no platform filter. Combines with other filter fields using AND.

型: string。任意。

Comma-separated topic IDs for the answer population. Matches any selected topic in the workspace. Omit or send an empty value for no topic filter. Combines with other filter fields using AND.

型: string。任意。

Comma-separated IDs of the brands that own the answer prompts. Matches any selected parent brand in the workspace; this does not select every brand mentioned in an answer. Omit for no parent-brand filter. Combines with other filter fields using AND.

型: string。任意。

Comma-separated prompt tag names, not tag IDs. Values are trimmed and lowercased. Matches any selected name. Omit or send an empty value for no tag filter. Combines with other filter fields using AND.

型: string。任意。

増えているsourceを期間比較します。

日次引用数を取得します。

競合gapをURL単位で調べます。

platform・topic・brand・tagのcomma-separated値は同じfield内でanyを満たし、異なるfield同士はANDです。brandは回答promptを所有するbrandで、任意の言及brandではありません。tagはIDでなく名前です。domain filterはsubdomainを登録可能domainへ正規化します。classification値はviewとscopeに合わせます。

認証し、対象Workspaceを選択して実行します。対象Workspaceへのアクセスが必要です。

orc sources では、次のグローバルオプションを使用できます。

各オプションの詳細と使用例は、グローバルオプションを参照してください。