--- language: en title: "meta-externalagent" description: "meta-externalagent identifies Meta’s crawler for training and direct indexing. Learn how to check specific rules without confusing it with other agents." canonical: "https://tropk.ai/en/resources/glossary/meta-externalagent" last_modified: "2026-10-12T16:11:15.189Z" --- # meta-externalagent Reviewed on 2026-10-06. ## What it means meta-externalagent is the identifier for Meta’s [crawler](/en/resources/glossary/crawler), intended for uses such as training foundational AI models and improving products through direct content [indexing](/en/resources/glossary/indexing). The operator’s documentation lists this purpose separately from research agents and agents for user-requested queries. The token lets you target [robots.txt](/en/resources/glossary/robots-txt) rules to this specific crawler. ## Why it matters for marketing Those managing an editorial archive need to decide which types of data collection they intend to allow. For marketing, this means translating that decision into a specific rule, without treating all Meta access as social sharing or search. Allowing meta-externalagent permits the [crawling](/en/resources/glossary/crawling) described by the operator, but does not let you measure how much a text will influence a model or product. ## Example in practice Suppose a publisher wants to exclude its interview directory from collection for training by meta-externalagent. The robots.txt file allows the rest of the site but does not mention this directory in the agent’s rule. The archive manager identifies the difference between the editorial policy and the file served and requests that a specific restriction be added for the interviews. ## How to apply or verify The person responsible for the site should check the robots.txt file for the interview host, the meta-externalagent token, and the restricted path. After the proposed change, they can compare new requests with the rule served. Meta says changes may take 24 hours due to caching. The User-Agent declares an identity; its text alone does not authenticate the connection’s origin. ## Important distinctions meta-externalagent is not [meta-webindexer](/en/resources/glossary/meta-webindexer), whose stated purpose involves [Meta AI](/en/resources/glossary/meta-ai) search, or [meta-externalfetcher](/en/resources/glossary/meta-externalfetcher), which fetches links requested by users. The latter’s robots.txt exemption should not be applied to the former. Blocking new collections does not demonstrate the retroactive removal of interviews already obtained; allowing access also does not guarantee indexing, [citation](/en/resources/glossary/ai-citation), or placement in responses. ## Sources and references - [Meta web crawlers]() — Meta ## Continue a leitura - [meta-externalfetcher]() - [meta-webindexer]() - [robots.txt]() ## About the author [Vinicius Pacheco]() Responsible for Tropk’s technical content, focusing on crawling, indexing, structured data, and agent access to content. --- [All pages in Markdown]()