methodology / AI Readiness / #44
AI plugin manifest (.well-known/ai-plugin.json)
#44 · Variable · Web Quality · weighted · AI Readiness · weight 0.7% · impl implemented · method v1.2.0
Web Quality factor
This factor is part of Web Quality — the weighted 0..100 score that sits above Web Standards. Its weight depends on what kind of site is being measured. Web Standards items take priority; this factor only enters the score once Web Standards passes.
- Base weight
- 0.2 applied to every site type unless overridden below
- Why this weight
- AI plugin manifest — niche emerging spec. Low weight until adoption tips.
Per-site-type overrides
| Site type | Weight | Δ vs base |
|---|---|---|
| Corporate / B2B | 0.3 | +0.1 |
| News / Publisher | 0.4 | +0.2 |
| SaaS / Product | 0.6 | +0.4 |
Site types not listed inherit the base weight.
What this means for your business
A small file you can publish lets AI tools talk to your site directly — checking availability, prices, or booking. Without it, you're missing out as customers shift from Googling to asking ChatGPT.
Plain title: A direct line for AI assistants to your business
What we measure
An AI plugin manifest tells AI agents how to interact with your site programmatically. It's an emerging standard that early adopters use to surface their services to LLM-based tools.
How to improve your score
Publish a minimal `ai-plugin.json` at `/.well-known/ai-plugin.json` describing your services, contact, and any public APIs.
Implementation
stale · v1 · seeded — no connector publish yet · source: freshcoat-discovery/src/connectors/seo-schema.ts:scoreAiPluginManifest
Detection method
GET <domain>/.well-known/ai-plugin.json. Present = pass. Absent = null (n/a) — the OpenAI Plugins spec was deprecated in 2024; treating its absence as a fail would penalise 99% of the web for a withdrawn convention.
Detection sources
- GET <domain>/.well-known/ai-plugin.json
Scoring bands · n_a_only ladder
| Score | Condition |
|---|---|
| 100 | manifest is present and parses as JSON |
| n/a | manifest is absent — spec is deprecated, absence isn't a quality signal |
Evidence-key dictionary
What every notes string the connector emits means.
Surfaces in the per-domain dossier evidence column.
ai_plugin_present- Manifest exists.
ai_plugin_spec_deprecated_or_unused- No manifest. The OpenAI Plugins spec was deprecated in 2024; n/a.
Applicability
Variable tier. Spec is deprecated upstream — kept as a positive signal only (presence rewarded; absence is n/a, not a fail). Apr 29 fix: was scoring 30 on every site, now correctly nulls out.
Changelog
- 2026-04-29 · seed Initial seed from MethodologyRegistry bootstrap.
Facts
Implementation notes
Single GET request to `/.well-known/ai-plugin.json`, parse and validate.
Scoring
Scoring formulas are versioned with the methodology. The current method (v1.2.0) maps raw measurements to pass, warn, fail. Factor weights determine how much each contributes to the composite — see the methodology index for the full table.
Cited by these standards
Standards in the Standards Library whose satisfiedBy requirement tree references this factor. Each link goes to the standard's full entry — methodology, scope, and the other factors it relies on.
Version history
| Version | Change | Date |
|---|---|---|
| v1.2.0 | Factor introduced. Status: live. Scoring impl: implemented. | 2026-04-25 |