/v1/oauth/introspectIntrospect an access token (RFC 7662)
Client-authenticated, and a client may only introspect its own tokens. `policy` is the ceiling the human approved; `effective_policy` is that intersected with the human's current workspace policy, resolved live on every call with no cache, so a demotion or an offboarding lands on the next call rather than at token expiry.
Request body
token*stringThe access token to introspect.
client_idstringclient_secretstringResponse · 200
active*booleanFalse for an unknown, expired, revoked, or foreign token — the four are deliberately indistinguishable.
scopestringPreset name(s) the grant started from. Not authority.
client_idstringsubstring · uuidThe human who consented.
workspace_idstring · uuidaudstringResource indicator the token is bound to.
expnumberExpiry, unix seconds.
policyarrayThe policy stored on the token — the ceiling the human approved at consent.
effective_policyarrayEffective authority: the token's policy intersected with the human's CURRENT workspace policy, recomputed on every call with no cache. An action the human's role loses disappears here without anyone revoking the token; a human who left the workspace makes the token report `active: false`. This is a report — the authorization decision itself re-evaluates both policies per (action, resource) cell.
Example
Request
POST https://api.platform-auth.productcraft.co/v1/oauth/introspect
Content-Type: application/json
{
"token": "string",
"client_id": "string",
"client_secret": "string"
}Response
{
"active": false,
"scope": "string",
"client_id": "string",
"sub": "00000000-0000-0000-0000-000000000000",
"workspace_id": "00000000-0000-0000-0000-000000000000",
"aud": "string",
"exp": 0,
"policy": [
{
"effect": "allow",
"actions": [
"social.read",
"social.list"
],
"resources": [
"*"
]
}
],
"effective_policy": [
{
"effect": "allow",
"actions": [
"social.read",
"social.list"
],
"resources": [
"*"
]
}
]
}