Implementations
Public record of conformant implementations of the AI Discovery Files specification.
This page enumerates the known conformant implementations. New implementations are added here when verified. See specification conventions for status definitions.
Last updated:
A specification matters in proportion to the implementations that depend on it. This page is the public record of implementations: who is publishing or consuming AI Discovery Files, what they support, and how each implementation maps to the conformance classes. The pattern follows the IETF tradition of implementation reports: a specification is taken more seriously when multiple independent implementations interoperate.
1. How to be listed
An implementation is listed here when:
- It produces, consumes, or validates one or more AI Discovery Files
- The behaviour can be verified by an independent party (for publisher tools: produce a file the AI Visibility Checker accepts; for consumer tools: produce output conforming to the validator-output schema)
- The implementer is happy for the implementation to be named publicly
To request a listing, contact AI Visibility via the support page. Self-declaration is acceptable; the maintainer reserves the right to verify the claim before listing. Listing here is not a certification mark; see claiming conformance for the rules on what publishers and tools MAY say about their conformance status.
2. Implementation role types
The AI Discovery Files ecosystem has four implementation role types. Most tools fill one role; some fill several.
- Publisher
- Generates one or more AI Discovery Files for a website. WordPress plugins, static-site generators, hand-edited templates, and the AI Discovery Files Service Pack all fall here.
- Validator
- Reads files at a site and produces a validation report. The output SHOULD conform to the validator-output schema. The AI Visibility Checker is the reference validator.
- Registry
- Maintains a public list of sites with AI Discovery Files, including computed conformance status. The AI Visible Directory is the current registry.
- Consumer
- An AI system, crawler, or downstream tool that reads AI Discovery Files and acts on them. See AI Consumer Guidance.
3. Current implementations
The table below summarises the implementations that the AI Discovery Files maintainer is aware of and has verified. Files-supported counts use the conformance class names from the conformance specification.
| Implementation | Role | Files supported | Conformance class | Status |
|---|---|---|---|---|
| AI Discovery Files (WordPress plugin) | Publisher | All 10 | Complete | Active |
| AI Visibility Checker | Validator | All 10 | n/a (reads, does not publish) | Active Reference implementation |
| AI Visible Directory | Registry + Validator | All 10 | n/a (reads + records) | Active |
| AI Discovery Files templates | Publisher (manual reference) | All 10 | Complete | Active |
| AI Discovery Files Service Pack | Publisher (managed) | All 10 | Complete | Active |
Third-party implementations are welcome and will be listed here when verified. As of the date above, no third-party implementations have been verified by the maintainer; the canonical specification, schemas, and examples are openly licensed (see licensing) and exist for any party to build against.
4. AI Discovery Files (WordPress plugin)
The AI Discovery Files WordPress plugin is a publisher implementation that generates all 10 AI Discovery Files from a WordPress site's existing settings, post content, and a plugin-specific settings panel.
Role: Publisher. Files supported: All 10. Conformance class: Complete.
The plugin maps WordPress site identity (blogname, blogdescription, the configured admin email, taxonomy terms, plus plugin-managed fields) to the corresponding entries in each AI Discovery File. Settings are exposed in the WordPress admin under Settings → AI Discovery Files. Files are served at the canonical root-level paths via WordPress's URL routing rather than written to the filesystem, so deployment-environment file permissions are not a concern.
The plugin is distributed on the WordPress.org plugin directory and licensed GPLv2-or-later (per WordPress.org plugin requirements). The plugin's own readme is the source of truth for current version and changelog; this page lists it as a registered implementation rather than mirroring its release notes.
5. AI Visibility Checker
The AI Visibility Checker is the validator implementation referenced throughout this specification as the reference implementation.
Role: Validator. Files supported: All 10. Output: conforms to the validator-output schema.
The checker fetches every AI Discovery File at the canonical root-level paths, parses each according to its specification, applies the cross-file consistency rules in the Interoperability Guide, executes the processing model to produce a normalised summary, and reports findings to the user. Output is rendered both as a human-readable HTML report and as machine-readable JSON conforming to validator-output.schema.json.
The reference-implementation status means the checker's behaviour, when the specification is ambiguous, is the de facto resolution. Where the checker's behaviour disagrees with the written specification, the written specification wins and the checker is the bug.
6. AI Visible Directory
The AI Visible Directory is the registry implementation. Publishers submit their domain, the directory validates the AI Discovery Files at that domain, and the result is recorded in a public listing with a computed conformance status.
Role: Registry + Validator. Files supported: All 10. Re-check cadence: Quarterly (automated) and on-demand (publisher-triggered).
The directory acts as a public ledger: it records who has been validated, what they were validated against, and when the last re-check ran. It is the canonical conformance registry for the specification; the formal role is documented at The Conformance Registry. Listings show the conformance class (Essential, Recommended, Complete) the publisher achieves; certification badges (when shipped) will be issued exclusively by the Directory.
7. AI Discovery Files templates
The AI Discovery Files templates are a publisher implementation for sites without WordPress or another generator. The templates are the canonical reference: a publisher who edits the placeholders and uploads the files produces a Complete-class implementation by hand.
Role: Publisher (manual reference). Files supported: All 10. Conformance class: Complete.
The templates double as documentation. The canonical structure of each file is encoded in the template; the placeholders are commented; the bundled README walks through the placement on the server. Anyone implementing AI Discovery Files for a non-WordPress site is expected to start from these templates.
8. AI Discovery Files Service Pack
The AI Discovery Files Service Pack is a managed publisher implementation: the customer commissions, 365i produces and deploys all 10 files. It exists for sites where the customer wants the work done but doesn't want to operate WordPress, edit templates by hand, or maintain the files themselves.
Role: Publisher (managed). Files supported: All 10. Conformance class: Complete on delivery; ongoing conformance is the publisher's responsibility unless covered by a maintenance agreement.
The service-pack implementation is the same technical work as the manual-template route; the difference is who performs the work, not the resulting files.
9. Third-party implementations
Third-party tools (validators, plugins, libraries) that consume or produce AI Discovery Files MAY claim conformance per the rules in the conformance specification. The maintainer encourages third-party implementations and will list verified ones on this page.
To get a third-party implementation listed:
- Build against the public specification, schemas, and registry
- If the tool is a validator, output conforming to the validator-output schema is the strongest evidence of conformance
- If the tool is a publisher, the output files should pass the AI Visibility Checker
- Open a request via support with a link to the tool and a sentence on what it does
The maintainer will not list private, paid-only, or unreleased tools without a way for an independent reviewer to verify the claim.
10. Licensing for implementers
The following licensing applies, and is repeated here so implementers have a single reference point:
- Specification text and HTML
- CC BY 4.0 (attribution required). Implementers MAY quote and reformat the specification; the canonical version remains the published one.
- JSON Schemas
- MIT. Implementers MAY embed, fork, or re-publish the schemas; attribution appreciated but not required.
- Canonical examples
- CC BY 4.0. Free to copy and adapt.
- Use of the name "AI Discovery Files"
- Permitted for accurate reference to the specification (e.g. "This plugin generates AI Discovery Files"). Not permitted in a way that implies endorsement or affiliation with AI Visibility / 365i without separate written permission.
- Conformance claims
- Self-declaration is permitted with the wording defined in the conformance specification. A formal certification mark is not currently issued; if one becomes available it will be a separate, opt-in process.
References
- Conformance specification: classes, claims, and conformance terminology.
- Validator-output JSON Schema: the contract validator implementations SHOULD honour.
- Specification registry (
registry.json): the machine-readable list of files, media types, and schema URLs that implementers build against. - Processing model: the algorithm a consumer or validator MUST follow.
- AI Consumer Guidance: how AI systems SHOULD act on the files.
- Relationship to other standards: where this specification sits in the wider landscape.
- RFC 2026: the IETF Internet standards process, source of the implementation-report pattern this page follows.