2026-03-13 01:35:34 +01:00
|
|
|
// Package oidc implements OIDC profile endpoints for KeyCape.
|
|
|
|
|
// Only profile-supported features are advertised — no implicit flow,
|
|
|
|
|
// no dynamic registration, no request objects.
|
|
|
|
|
package oidc
|
|
|
|
|
|
|
|
|
|
import (
|
|
|
|
|
"encoding/json"
|
|
|
|
|
"net/http"
|
Reconcile the canonical model and discovery with the runtime
Closes gap G02 of the scope assessment for the client-registration and discovery
surface. spec/canonical-model.yaml and domain/model.go both claimed to be the
source of truth and disagreed: the spec restricted grants to authorization_code,
required redirect URIs of every client, and omitted the audience, service
subject, tenant, role, MFA and handoff fields the runtime reads.
The durable part is the link, not the edit. A two-way conformance test compares
the spec against the Go model by reflection and fails when a runtime field has
no spec entry or a spec entry is not read by the runtime, the latter unless
marked runtime: false. It found drift beyond the assessment's list on its first
run -- User.tenant was undeclared -- which is the argument for the check over a
one-time reconciliation.
Discovery now advertises the core profile claims that appear on every token and
derives scopes_supported from the registered clients rather than a fixed list.
The Go model is stated as the runtime authority and the YAML as the reviewed
contract, in both files. This covers client registration and discovery, not
schema enforcement in general, which remains G06.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01NV9oijZukGyGbRQGGKnK4P
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 713576@bnt-lap001
Assistant-Session: 384c511d-9bce-4cb8-a676-2aef6c0c8df6
2026-09-07 00:22:49 +02:00
|
|
|
"sort"
|
|
|
|
|
|
|
|
|
|
"keycape/internal/domain"
|
2026-03-13 01:35:34 +01:00
|
|
|
)
|
|
|
|
|
|
|
|
|
|
// DiscoveryConfig holds the issuer and endpoint URLs for the discovery document.
|
|
|
|
|
// UserinfoEndpoint is optional; if empty it is omitted from the document.
|
|
|
|
|
type DiscoveryConfig struct {
|
|
|
|
|
Issuer string // e.g. "https://auth.netkingdom.local"
|
|
|
|
|
AuthorizationEndpoint string
|
|
|
|
|
TokenEndpoint string
|
|
|
|
|
JWKSUri string
|
|
|
|
|
UserinfoEndpoint string // optional, empty = not advertised
|
2026-08-16 01:05:27 +02:00
|
|
|
EndSessionEndpoint string // optional, empty = not advertised
|
Reconcile the canonical model and discovery with the runtime
Closes gap G02 of the scope assessment for the client-registration and discovery
surface. spec/canonical-model.yaml and domain/model.go both claimed to be the
source of truth and disagreed: the spec restricted grants to authorization_code,
required redirect URIs of every client, and omitted the audience, service
subject, tenant, role, MFA and handoff fields the runtime reads.
The durable part is the link, not the edit. A two-way conformance test compares
the spec against the Go model by reflection and fails when a runtime field has
no spec entry or a spec entry is not read by the runtime, the latter unless
marked runtime: false. It found drift beyond the assessment's list on its first
run -- User.tenant was undeclared -- which is the argument for the check over a
one-time reconciliation.
Discovery now advertises the core profile claims that appear on every token and
derives scopes_supported from the registered clients rather than a fixed list.
The Go model is stated as the runtime authority and the YAML as the reviewed
contract, in both files. This covers client registration and discovery, not
schema enforcement in general, which remains G06.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01NV9oijZukGyGbRQGGKnK4P
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 713576@bnt-lap001
Assistant-Session: 384c511d-9bce-4cb8-a676-2aef6c0c8df6
2026-09-07 00:22:49 +02:00
|
|
|
// Clients supplies the registered clients so scopes_supported reflects the
|
|
|
|
|
// scopes this deployment can actually grant, rather than a fixed list that
|
|
|
|
|
// omits every configured resource scope (KEY-WP-0017-T03). Nil or empty
|
|
|
|
|
// leaves only the baseline OIDC scopes advertised.
|
|
|
|
|
Clients map[string]*domain.Client
|
2026-03-13 01:35:34 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// discoveryDocument is the JSON shape of /.well-known/openid-configuration.
|
|
|
|
|
// Fields are ordered to match common OIDC implementations for readability.
|
|
|
|
|
// registration_endpoint is intentionally absent — no dynamic client registration.
|
|
|
|
|
type discoveryDocument struct {
|
|
|
|
|
Issuer string `json:"issuer"`
|
|
|
|
|
AuthorizationEndpoint string `json:"authorization_endpoint"`
|
|
|
|
|
TokenEndpoint string `json:"token_endpoint"`
|
|
|
|
|
JWKSUri string `json:"jwks_uri"`
|
|
|
|
|
UserinfoEndpoint string `json:"userinfo_endpoint,omitempty"`
|
2026-08-16 01:05:27 +02:00
|
|
|
EndSessionEndpoint string `json:"end_session_endpoint,omitempty"`
|
2026-03-13 01:35:34 +01:00
|
|
|
ResponseTypesSupported []string `json:"response_types_supported"`
|
|
|
|
|
GrantTypesSupported []string `json:"grant_types_supported"`
|
|
|
|
|
CodeChallengeMethodsSupported []string `json:"code_challenge_methods_supported"`
|
|
|
|
|
IDTokenSigningAlgValuesSupported []string `json:"id_token_signing_alg_values_supported"`
|
|
|
|
|
ScopesSupported []string `json:"scopes_supported"`
|
|
|
|
|
TokenEndpointAuthMethodsSupported []string `json:"token_endpoint_auth_methods_supported"`
|
|
|
|
|
ClaimsSupported []string `json:"claims_supported"`
|
|
|
|
|
SubjectTypesSupported []string `json:"subject_types_supported"`
|
|
|
|
|
RequestParameterSupported bool `json:"request_parameter_supported"`
|
|
|
|
|
ClaimsParameterSupported bool `json:"claims_parameter_supported"`
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// discoveryHandler implements http.Handler for GET /.well-known/openid-configuration.
|
|
|
|
|
type discoveryHandler struct {
|
|
|
|
|
doc []byte
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// NewDiscoveryHandler returns an http.Handler that serves the OIDC discovery document.
|
|
|
|
|
// The document is pre-serialised at construction time so every request is a cheap copy.
|
|
|
|
|
func NewDiscoveryHandler(cfg DiscoveryConfig) http.Handler {
|
|
|
|
|
d := discoveryDocument{
|
|
|
|
|
Issuer: cfg.Issuer,
|
|
|
|
|
AuthorizationEndpoint: cfg.AuthorizationEndpoint,
|
|
|
|
|
TokenEndpoint: cfg.TokenEndpoint,
|
|
|
|
|
JWKSUri: cfg.JWKSUri,
|
|
|
|
|
UserinfoEndpoint: cfg.UserinfoEndpoint,
|
2026-08-16 01:05:27 +02:00
|
|
|
EndSessionEndpoint: cfg.EndSessionEndpoint,
|
2026-03-13 01:35:34 +01:00
|
|
|
|
|
|
|
|
// Profile-locked values — not negotiable.
|
|
|
|
|
ResponseTypesSupported: []string{"code"},
|
2026-07-27 20:03:07 +02:00
|
|
|
GrantTypesSupported: []string{"authorization_code", "client_credentials"},
|
2026-03-13 01:35:34 +01:00
|
|
|
CodeChallengeMethodsSupported: []string{"S256"},
|
|
|
|
|
IDTokenSigningAlgValuesSupported: []string{"RS256"},
|
Reconcile the canonical model and discovery with the runtime
Closes gap G02 of the scope assessment for the client-registration and discovery
surface. spec/canonical-model.yaml and domain/model.go both claimed to be the
source of truth and disagreed: the spec restricted grants to authorization_code,
required redirect URIs of every client, and omitted the audience, service
subject, tenant, role, MFA and handoff fields the runtime reads.
The durable part is the link, not the edit. A two-way conformance test compares
the spec against the Go model by reflection and fails when a runtime field has
no spec entry or a spec entry is not read by the runtime, the latter unless
marked runtime: false. It found drift beyond the assessment's list on its first
run -- User.tenant was undeclared -- which is the argument for the check over a
one-time reconciliation.
Discovery now advertises the core profile claims that appear on every token and
derives scopes_supported from the registered clients rather than a fixed list.
The Go model is stated as the runtime authority and the YAML as the reviewed
contract, in both files. This covers client registration and discovery, not
schema enforcement in general, which remains G06.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01NV9oijZukGyGbRQGGKnK4P
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 713576@bnt-lap001
Assistant-Session: 384c511d-9bce-4cb8-a676-2aef6c0c8df6
2026-09-07 00:22:49 +02:00
|
|
|
ScopesSupported: supportedScopes(cfg.Clients),
|
2026-07-27 20:03:07 +02:00
|
|
|
TokenEndpointAuthMethodsSupported: []string{"client_secret_basic", "none"},
|
Reconcile the canonical model and discovery with the runtime
Closes gap G02 of the scope assessment for the client-registration and discovery
surface. spec/canonical-model.yaml and domain/model.go both claimed to be the
source of truth and disagreed: the spec restricted grants to authorization_code,
required redirect URIs of every client, and omitted the audience, service
subject, tenant, role, MFA and handoff fields the runtime reads.
The durable part is the link, not the edit. A two-way conformance test compares
the spec against the Go model by reflection and fails when a runtime field has
no spec entry or a spec entry is not read by the runtime, the latter unless
marked runtime: false. It found drift beyond the assessment's list on its first
run -- User.tenant was undeclared -- which is the argument for the check over a
one-time reconciliation.
Discovery now advertises the core profile claims that appear on every token and
derives scopes_supported from the registered clients rather than a fixed list.
The Go model is stated as the runtime authority and the YAML as the reviewed
contract, in both files. This covers client registration and discovery, not
schema enforcement in general, which remains G06.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01NV9oijZukGyGbRQGGKnK4P
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 713576@bnt-lap001
Assistant-Session: 384c511d-9bce-4cb8-a676-2aef6c0c8df6
2026-09-07 00:22:49 +02:00
|
|
|
// Core profile claims are emitted on every token, so omitting one here
|
|
|
|
|
// is a contract defect rather than missing optional metadata. The
|
|
|
|
|
// scope-gated human claims and the optional cached tenant_roles claim
|
|
|
|
|
// follow them.
|
2026-03-13 01:35:34 +01:00
|
|
|
ClaimsSupported: []string{
|
|
|
|
|
"sub", "iss", "aud", "exp", "iat",
|
Carry the tenant claim's provenance, and correct a guard the ruling voided
GH-DEC-2026-013 §5 is a finding nobody asked for and ours to implement: tenant is
a bare string, so a consumer cannot tell a zone the directory asserted about the
person from one a registration supplied about the client they came through.
approval-engine exact-matches that string while its contract reads as though it
relies on the first -- the check is sound and the property a reader infers from
it is absent. gate-house named the property and left the mechanism to us.
Every token now carries tenant_source beside tenant: directory, registration or
default. Advertised in claims_supported, and asserted at the token level on both
grants rather than only in the resolution function, since the claim a consumer
reads is the thing under obligation.
Three values where the ruling names two, which is the judgement here. Labelling
an unasserted profile default as directory would reproduce the same defect one
level down -- a consumer reading an assertion the identity layer never made. The
ruling cites GH-DEC-2026-011 §3 on unknown versus absent for the case it
examined; the same rule applies to our own fallback. The agreement case resolves
to directory deliberately: if a registration declares the zone the directory also
assigned, the directory did assert it, and reporting the weaker source would
understate what is known.
Also corrects the guard shipped in 5f516a0. Its failure message offered two ways
out of adding dynamic registration, and condition (b) voids the second: admitting
dynamic registration voids the registration-bound shape that day, whatever state
the adapter is in. The message named an inadmissible resolution in the exact
place someone would read it while making that change.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01NV9oijZukGyGbRQGGKnK4P
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 713576@bnt-lap001
Assistant-Session: 384c511d-9bce-4cb8-a676-2aef6c0c8df6
2026-09-10 07:58:29 +02:00
|
|
|
"tenant", "tenant_source", "principal_type", "roles", "groups", "assurance", "scope",
|
Reconcile the canonical model and discovery with the runtime
Closes gap G02 of the scope assessment for the client-registration and discovery
surface. spec/canonical-model.yaml and domain/model.go both claimed to be the
source of truth and disagreed: the spec restricted grants to authorization_code,
required redirect URIs of every client, and omitted the audience, service
subject, tenant, role, MFA and handoff fields the runtime reads.
The durable part is the link, not the edit. A two-way conformance test compares
the spec against the Go model by reflection and fails when a runtime field has
no spec entry or a spec entry is not read by the runtime, the latter unless
marked runtime: false. It found drift beyond the assessment's list on its first
run -- User.tenant was undeclared -- which is the argument for the check over a
one-time reconciliation.
Discovery now advertises the core profile claims that appear on every token and
derives scopes_supported from the registered clients rather than a fixed list.
The Go model is stated as the runtime authority and the YAML as the reviewed
contract, in both files. This covers client registration and discovery, not
schema enforcement in general, which remains G06.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01NV9oijZukGyGbRQGGKnK4P
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 713576@bnt-lap001
Assistant-Session: 384c511d-9bce-4cb8-a676-2aef6c0c8df6
2026-09-07 00:22:49 +02:00
|
|
|
"nonce", "preferred_username", "email", "name", "tenant_roles",
|
2026-03-13 01:35:34 +01:00
|
|
|
},
|
|
|
|
|
SubjectTypesSupported: []string{"public"},
|
|
|
|
|
RequestParameterSupported: false,
|
|
|
|
|
ClaimsParameterSupported: false,
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
b, err := json.Marshal(d)
|
|
|
|
|
if err != nil {
|
|
|
|
|
// This can only fail if the struct contains un-marshallable types, which it does not.
|
|
|
|
|
panic("oidc: failed to marshal discovery document: " + err.Error())
|
|
|
|
|
}
|
|
|
|
|
return &discoveryHandler{doc: b}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
func (h *discoveryHandler) ServeHTTP(w http.ResponseWriter, _ *http.Request) {
|
|
|
|
|
w.Header().Set("Content-Type", "application/json")
|
|
|
|
|
w.Header().Set("Cache-Control", "max-age=3600")
|
|
|
|
|
w.WriteHeader(http.StatusOK)
|
|
|
|
|
_, _ = w.Write(h.doc)
|
|
|
|
|
}
|
Reconcile the canonical model and discovery with the runtime
Closes gap G02 of the scope assessment for the client-registration and discovery
surface. spec/canonical-model.yaml and domain/model.go both claimed to be the
source of truth and disagreed: the spec restricted grants to authorization_code,
required redirect URIs of every client, and omitted the audience, service
subject, tenant, role, MFA and handoff fields the runtime reads.
The durable part is the link, not the edit. A two-way conformance test compares
the spec against the Go model by reflection and fails when a runtime field has
no spec entry or a spec entry is not read by the runtime, the latter unless
marked runtime: false. It found drift beyond the assessment's list on its first
run -- User.tenant was undeclared -- which is the argument for the check over a
one-time reconciliation.
Discovery now advertises the core profile claims that appear on every token and
derives scopes_supported from the registered clients rather than a fixed list.
The Go model is stated as the runtime authority and the YAML as the reviewed
contract, in both files. This covers client registration and discovery, not
schema enforcement in general, which remains G06.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01NV9oijZukGyGbRQGGKnK4P
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 713576@bnt-lap001
Assistant-Session: 384c511d-9bce-4cb8-a676-2aef6c0c8df6
2026-09-07 00:22:49 +02:00
|
|
|
|
|
|
|
|
// supportedScopes returns the baseline OIDC scopes plus every scope any
|
|
|
|
|
// registered client is allowed to request, sorted for a stable document. The
|
|
|
|
|
// baseline is advertised even with no clients configured, since it describes the
|
|
|
|
|
// profile rather than a particular deployment.
|
|
|
|
|
func supportedScopes(clients map[string]*domain.Client) []string {
|
|
|
|
|
seen := map[string]bool{}
|
|
|
|
|
scopes := []string{}
|
|
|
|
|
for _, scope := range []string{"openid", "profile", "email", "groups"} {
|
|
|
|
|
seen[scope] = true
|
|
|
|
|
scopes = append(scopes, scope)
|
|
|
|
|
}
|
|
|
|
|
extra := []string{}
|
|
|
|
|
for _, client := range clients {
|
|
|
|
|
for _, scope := range client.AllowedScopes {
|
|
|
|
|
if scope == "" || seen[scope] {
|
|
|
|
|
continue
|
|
|
|
|
}
|
|
|
|
|
seen[scope] = true
|
|
|
|
|
extra = append(extra, scope)
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
sort.Strings(extra)
|
|
|
|
|
return append(scopes, extra...)
|
|
|
|
|
}
|