Platform llms.txt vs Brand llms.txt
> Platform llms.txt and brand llms.txt are two distinct types of AI-readable files generated by FeedBridge — the platform llms.txt describes FeedBridge itself as a software platform, while brand llms.txt files are generated per merchant brand and describe each merchant's identity, product range, and key content — serving different AI discoverability purposes for different subjects.
---
What the Distinction Is
The `llms.txt` standard defines a format, not a scope. Any organisation — a software platform, a retailer, a media publication — can publish a `llms.txt` file at their root domain to give AI language models a curated, structured guide to their most important content. The file format is the same; what differs is the subject matter the file describes and the audience whose AI discoverability it serves.
In FeedBridge's implementation, this distinction creates two parallel types of llms.txt assets:
- Platform llms.txt — a file describing FeedBridge as a platform: its features, integrations, capabilities, and documentation. This file serves FeedBridge's own AI discoverability — ensuring that AI assistants and research agents that are asked about FeedBridge can access accurate, structured, current information about what the platform does.
- Brand llms.txt — per-brand files, one generated for each merchant brand on FeedBridge, describing that specific brand: its identity, product range, target audience, and key content pages. These files serve each merchant brand's AI discoverability — ensuring that AI assistants asked about that brand can access accurate, merchant-curated information rather than reconstructed web crawl fragments.
---
Platform llms.txt: What It Is and Who It Serves
The FeedBridge platform `llms.txt` is an AI-readable file describing FeedBridge as a B2B SaaS product. It is maintained at the FeedBridge platform level — not generated per brand — and serves as the authoritative AI-readable reference for FeedBridge's platform identity. [file:4]
Subject
The platform `llms.txt` describes:
- What FeedBridge is (a B2B SaaS platform for AI commerce readiness)
- What it does (automates product feed creation, AI enrichment, and hosting)
- Who it serves (e-commerce brands, agencies, marketplace sellers)
- What its key features are (AI Readiness Score, ACP/UCP feed generation, enrichment pipeline, platform intelligence)
- Where its key documentation and public pages are
Audience
The primary audience for the platform `llms.txt` is:
- AI assistants asked by potential customers about FeedBridge's capabilities ("what does FeedBridge do?", "how does FeedBridge compare to other product feed tools?")
- AI research agents evaluating e-commerce technology providers
- AI systems building brand or vendor knowledge bases that include FeedBridge as a referenced platform
Platform llms-full.txt
FeedBridge also maintains a `llms-full.txt` — a comprehensive long-form documentation file covering FeedBridge's complete feature set in detail. Where `llms.txt` is a compact summary, `llms-full.txt` is a deep-reference file: 461 lines of structured documentation covering every major capability area, live features, known gaps, and integration details. [file:4]
The `llms-full.txt` serves AI systems that need comprehensive platform information — a research agent evaluating FeedBridge for a detailed comparison, or an AI assistant building a knowledge file for a Custom GPT that needs full platform context. The compact `llms.txt` serves systems that need orientation; `llms-full.txt` serves systems that need depth.
---
Brand llms.txt: What It Is and Who It Serves
A brand `llms.txt` is an AI-readable file describing a specific merchant brand. FeedBridge generates one per brand — each unique to that merchant's identity, product scope, and content structure. [file:4]
Subject
Each brand `llms.txt` describes:
- The brand name and a neutral, factual summary of what the brand is and sells
- The brand's product categories and core product range
- The target audience and buyer persona the brand serves
- Links to key product pages, category pages, brand story, and FAQs
- Any structured product or policy documentation that provides relevant brand context
Audience
The primary audience for a brand `llms.txt` is:
- AI shopping assistants asked by buyers about the brand ("what does [Brand] sell?", "is [Brand] right for me?", "what are [Brand]'s most popular products?")
- AI agents evaluating brands within a product category for recommendation
- AI chat tools (Custom GPTs, Gemini Gems, WhatsApp bots) built on or referencing the brand's product catalog
Why Per-Brand Generation Matters
The per-brand structure is essential because AI discoverability for a merchant brand requires brand-specific content. A generic `llms.txt` template that describes a brand's category without specifics does not give AI models useful differentiation information — it tells the model the brand sells electronics or apparel or beauty products, but not what makes that brand's specific product range distinct or who it is designed for.
FeedBridge's brand `llms.txt` generation draws on the merchant's catalog data and brand profile in FeedBridge, producing a file that reflects the actual brand identity rather than a generic category description. This brand-specific accuracy is what enables AI assistants to make correct brand attributions when responding to buyer queries.
---
Side-by-Side Comparison
| Dimension | Platform llms.txt | Brand llms.txt | |---|---|---| | Subject | FeedBridge as a SaaS platform | Individual merchant brand | | Scope | Platform features, integrations, docs | Brand identity, products, audience, pages | | Generated how | Maintained by FeedBridge platform team | Generated per brand from merchant catalog data | | Instance count | One (plus llms-full.txt) | One per merchant brand on FeedBridge | | Primary AI audience | Research agents, vendor evaluators, platform-query AI assistants | Shopping assistants, product recommendation agents, brand-query AI assistants | | Hosted where | FeedBridge platform domain | Per-brand hosted URL (FeedBridge infrastructure) | | Content depth | Compact (llms.txt) + deep (llms-full.txt) | Single structured file per brand | | FeedBridge status | Live | Live |
---
How They Complement Each Other
Platform llms.txt and brand llms.txt operate at different levels of the AI discoverability stack and address different AI query types — which means they complement rather than duplicate each other.
Platform level handles the meta-context: when an AI agent or buyer asks about FeedBridge as a platform — "how does FeedBridge work?", "what formats does FeedBridge support?", "does FeedBridge integrate with Shopify?" — the platform llms.txt provides the structured context the model needs to answer accurately. Without it, AI systems reconstruct FeedBridge's capabilities from web crawl data, product pages, and third-party references — potentially returning outdated, incomplete, or inaccurate information.
Brand level handles the product-context: when a buyer asks an AI shopping assistant about a specific merchant brand — "what does [Brand] sell?", "is this brand good for sensitive skin?", "does [Brand] have products for professional use?" — the brand llms.txt provides the structured context the model needs to accurately represent that brand. Without it, the AI reconstructs brand identity from product pages, review sites, and cached content — with the same risks of inaccuracy and incompleteness.
For merchants who deploy AI assistants and chatbots using FeedBridge's AI assistant builder tools (Custom GPT Builder, Gemini Gem Builder, WhatsApp Bot Kit), both levels of llms.txt contribute to accuracy:
- The brand `llms.txt` informs the assistant's understanding of the specific merchant brand
- The platform `llms.txt` and `llms-full.txt` inform the system's understanding of FeedBridge's infrastructure when building system instructions and knowledge files for the assistant
Agency Context: Managing Both Levels
For agencies managing multiple merchant brands on FeedBridge, the two-level llms.txt structure creates a clear operational model:
- Platform llms.txt: Maintained by FeedBridge — agencies do not need to manage or update it. It is the platform's responsibility to keep the platform-level AI-readable documentation current.
- Brand llms.txt: Generated per client brand from FeedBridge's catalog data — agencies benefit from ensuring each client's catalog and brand profile in FeedBridge is current and complete, as this data feeds the brand llms.txt generation. Each client brand's llms.txt is a distinct file; updating one client's data does not affect others.
---
What Neither File Does
Understanding the limits of both llms.txt types prevents over-claiming:
- Neither file controls AI training data. `llms.txt` operates at inference time — when an AI model is answering a query and retrieving current information. It does not influence what the model learned during its pre-training phase. Brands that were poorly represented or absent in pre-training data cannot correct that through `llms.txt` alone. [web:81]
- Neither file guarantees AI citation or recommendation. AI assistants select sources and references based on multiple relevance and authority signals. A well-constructed `llms.txt` improves the accuracy of AI representation when the brand is referenced, but does not guarantee the brand will be referenced or recommended in response to any specific query.
- Neither file replaces product feed data. Platform and brand llms.txt files operate at the brand and platform identity layer. Product-level data — individual SKU attributes, pricing, availability, structured variants, trust signals — is provided through product feed formats (ACP JSON-LD, GMC CSV) and protocol endpoints (UCP Interactive Protocol). Both layers are necessary for complete AI commerce readiness; llms.txt alone does not enable product-level AI transactions.
Implementation Checklist
For merchants and agencies implementing llms.txt as part of an AI Shopping SEO strategy on FeedBridge:
- [ ] Confirm brand llms.txt is generated — verify that the per-brand llms.txt file has been generated for each brand in FeedBridge and is accessible at its hosted URL
- [ ] Verify the hosted URL is accessible to AI crawlers — confirm the URL returns the file without authentication requirements, so AI platform crawlers can index it
- [ ] Ensure catalog and brand profile data is current — since brand llms.txt content is generated from FeedBridge catalog data, the quality of the file reflects the quality of the underlying brand and product data; keep this current
- [ ] Reference platform llms.txt for system instructions — when building Custom GPT, Gemini Gem, or WhatsApp Bot knowledge files using FeedBridge's AI assistant builder tools, reference the platform llms-full.txt URL for comprehensive platform context
- [ ] Pair llms.txt with schema code generation — for complete AI discoverability coverage, combine llms.txt with FeedBridge's JSON-LD schema markup generation, which operates at the individual product and organisation data layer
Why It Matters for Merchants
For a merchant on FeedBridge, the practical value of brand llms.txt is that it exists and is generated automatically — it is not a manual content production task. The merchant's investment in maintaining an accurate catalog and brand profile in FeedBridge produces a brand `llms.txt` as an output, which then serves the brand's AI discoverability across every AI platform whose crawlers index it.
For agencies, the value is that every client brand managed on FeedBridge has a generated `llms.txt` asset — an AI-readable brand identity file at a stable hosted URL — that contributes to that client's AI Shopping SEO posture without requiring manual content creation per brand.
The combination of brand-level llms.txt (for brand identity), product feed data (for individual product attributes), and schema markup (for page-level structured data) represents the three-layer AI discoverability stack that gives merchants the best available infrastructure for accurate AI representation across the current landscape of AI shopping surfaces.
---
FeedBridge Relevance
FeedBridge generates both types as live features: per-brand `llms.txt` files for each merchant brand, and platform-level `llms.txt` and `llms-full.txt` for FeedBridge itself. Brand `llms.txt` files are hosted at stable per-brand URLs generated from the merchant's FeedBridge catalog and brand profile data. The platform `llms-full.txt` is a 461-line comprehensive documentation file. Both are part of FeedBridge's AI Shopping SEO capability set alongside schema code generation, voice SEO content management, and blog content hubs. [file:4]
---
Frequently Asked Questions
Q: Do merchants need to write their brand llms.txt manually? A: No. FeedBridge generates per-brand llms.txt files automatically from the merchant's catalog and brand profile data in the platform. Merchants do not need to write or maintain the file manually — it is an output of FeedBridge's AI Shopping SEO pipeline. The quality of the generated file reflects the quality and completeness of the underlying catalog and brand data in FeedBridge.
Q: Can a merchant customise the content of their brand llms.txt? A: The brand llms.txt is generated from FeedBridge catalog and brand profile data. Merchants can influence the content by keeping their catalog data, product descriptions, and brand profile current and complete in FeedBridge, as this data feeds the generation pipeline. The specific mechanism for directly editing the generated llms.txt content is a platform capability question beyond what is documented in the current capabilities report.
Q: Is the platform llms-full.txt different from the platform llms.txt? A: Yes. The platform `llms.txt` is a compact summary of FeedBridge's capabilities — designed for AI systems that need orientation about what FeedBridge is and does. The `llms-full.txt` is a comprehensive, 461-line long-form documentation file covering FeedBridge's full feature set in structured detail. The full version is intended for AI systems that need depth — such as a knowledge file for a Custom GPT built around FeedBridge's platform. [file:4]
Q: Should a brand have both a brand llms.txt and a schema.org JSON-LD implementation? A: Yes — these serve different but complementary purposes. The brand `llms.txt` gives AI models a structured narrative guide to the brand's content and identity, operating at the brand level. JSON-LD schema markup encodes structured machine-readable data at the individual page and product level. Both are part of FeedBridge's AI Shopping SEO capability set, and both contribute to AI discoverability — at different granularities and for different AI consumption patterns.
Q: Does the platform llms.txt benefit merchant brands directly? A: The platform `llms.txt` benefits merchants indirectly by ensuring AI assistants can accurately describe FeedBridge's capabilities when evaluating it as a platform — useful for merchants and agencies investigating FeedBridge, and for AI systems comparing product feed and AI commerce readiness tools. It does not directly serve any individual merchant brand's product discoverability; that is the role of the per-brand `llms.txt`.
---
Related Topics
Parent hub: AI Shopping SEO — llms.txt
Related concepts:
- Brand llms.txt for AI Discoverability
- llms-full.txt and Long-Form Documentation
- Schema Code Generation for Commerce Pages
- Taxonomy Normalization for Product Data
Breadcrumb:
---
Source Documentation
| Claim | Source | Source Class | Reference | |---|---|---|---| | Brand llms.txt: per-brand AI-readable brand files — live feature | FeedBridge Platform Capabilities April 2026 v2.0 | T1 – FeedBridge Internal | FeedBridge-Platform-Capabilities-April2026.md | | Platform llms.txt and llms-full.txt (461 lines) — live features | FeedBridge Platform Capabilities April 2026 v2.0 | T1 – FeedBridge Internal | FeedBridge-Platform-Capabilities-April2026.md | | Schema Code Generator, Voice SEO, Blog Content Hub — live AI Shopping SEO features | FeedBridge Platform Capabilities April 2026 v2.0 | T1 – FeedBridge Internal | FeedBridge-Platform-Capabilities-April2026.md | | Custom GPT Builder, Gemini Gem Builder, WhatsApp Bot Kit — live AI assistant features | FeedBridge Platform Capabilities April 2026 v2.0 | T1 – FeedBridge Internal | FeedBridge-Platform-Capabilities-April2026.md | | llms.txt operates at inference time, not training time; does not influence pre-training data | What is llms.txt? How It Affects AI Visibility — Visiblie | T2 – Ecosystem | visiblie.com/blog/what-is-llms-txt |