Error codes
Single sign-on, chat apps and connectors
Each code of this family: what it means, its most frequent cause, what to do and who can do it.
sso_domain_unverified
- What it means
- Single sign-on cannot be required for a domain the workspace has not verified it owns.
- Most frequent cause
- The domain’s DNS TXT record was not verified yet.
- What to do
- Add the TXT record Settings › Access shows to the domain’s DNS, verify it, then require single sign-on.
- Who can do it
- An owner or admin of the workspace · Open the screen
sso_provider_required
- What it means
- Single sign-on cannot be required while no identity provider is enabled for the workspace.
- Most frequent cause
- No provider configured yet, or the one configured is turned off.
- What to do
- Configure the identity provider in Settings › Access, test a sign-in with it, then require single sign-on.
- Who can do it
- An owner or admin of the workspace · Open the screen
sso_type_change
- What it means
- This workspace already has a SAML identity provider: an OIDC provider cannot be saved beside it.
- Most frequent cause
- A switch from SAML to OIDC without removing the SAML provider first.
- What to do
- Remove the SAML provider in Settings › Access, then configure OIDC.
- Who can do it
- An owner or admin of the workspace · Open the screen
oidc_client_secret_required
- What it means
- An OIDC identity provider is saved with its client secret.
- Most frequent cause
- The client secret left empty for a new provider.
- What to do
- Paste the client secret your identity provider gave for this application, then save.
- Who can do it
- An owner or admin of the workspace · Open the screen
oidc_issuer_invalid
- What it means
- The issuer is not a public https address an identity provider can use, so the provider was not saved.
- Most frequent cause
- A typo, an http address, or an address with a query, a fragment or credentials in it.
- What to do
- Enter the issuer exactly as your identity provider publishes it — https, no query — then save.
- Who can do it
- An owner or admin of the workspace · Open the screen
oidc_issuer_multi_tenant
- What it means
- This issuer serves many organisations, so it cannot sign in the members of one workspace.
- Most frequent cause
- A shared Microsoft Entra address (common or organizations) instead of your own tenant’s.
- What to do
- Use your own tenant’s issuer address, with its tenant id, then save.
- Who can do it
- An owner or admin of the workspace · Open the screen
oidc_discovery_failed
- What it means
- The identity provider’s discovery document could not be read, so the provider was not saved.
- Most frequent cause
- A wrong issuer address, or a provider Azoth cannot reach.
- What to do
- Check the issuer address, and that its /.well-known/openid-configuration answers, then save again.
- Who can do it
- An owner or admin of the workspace · Open the screen
oidc_discovery_mismatch
- What it means
- The identity provider’s discovery document names another issuer, or lacks an endpoint, so the provider was not saved.
- Most frequent cause
- An issuer entered with another path or ending than the provider’s own.
- What to do
- Enter the issuer exactly as its discovery document names it, then save.
- Who can do it
- An owner or admin of the workspace · Open the screen
sso_too_many_attempts
- What it means
- Too many single sign-on returns came from this network address in a minute, so this one was not completed.
- Most frequent cause
- Repeated sign-in attempts, or many people signing in at once behind one network address.
- What to do
- Wait a minute, then start the sign-in again.
- Who can do it
- Any member of the workspace
sso_attempt_expired
- What it means
- The sign-in that came back from the identity provider is not one this browser started, or it took too long, so no session was opened.
- Most frequent cause
- Too long spent at the identity provider, a sign-in started in another browser or tab, or a return link opened a second time.
- What to do
- Start the sign-in again from the sign-in page, in the same browser, and finish it without delay.
- Who can do it
- Any member of the workspace
sso_idp_refused
- What it means
- The identity provider came back without signing the person in, so no session was opened.
- Most frequent cause
- The sign-in was cancelled or consent refused, or the account is not allowed to use the Azoth application at the identity provider.
- What to do
- Ask your identity provider’s administrator to give your account access to the Azoth application, then sign in again.
- Who can do it
- An owner or admin of the workspace
sso_provider_misconfigured
- What it means
- The workspace’s identity provider settings cannot be used to sign in, so no session was opened.
- Most frequent cause
- An issuer Azoth no longer accepts — a shared Microsoft tenant, an address that is not public https — or a discovery document that names another issuer than the one saved.
- What to do
- An owner or admin corrects the identity provider in Settings › Access — the issuer exactly as the provider publishes it, your own tenant’s — then signs in again.
- Who can do it
- An owner or admin of the workspace · Open the screen
sso_token_exchange_failed
- What it means
- The identity provider refused to complete the sign-in for Azoth, so no session was opened.
- Most frequent cause
- A client secret changed or expired at the identity provider, a client ID that does not match, or Azoth’s redirect address not registered there.
- What to do
- An owner or admin checks the client ID and secret in Settings › Access, and that the redirect address shown there is registered at the identity provider.
- Who can do it
- An owner or admin of the workspace · Open the screen
sso_id_token_invalid
- What it means
- The identity the provider sent could not be verified, so no session was opened.
- Most frequent cause
- A token signed with an algorithm Azoth does not accept or by a key the provider does not publish, issued for another application, or already expired.
- What to do
- An owner or admin checks the identity provider’s token signing (RS256, PS256 or ES256) and the client ID in Settings › Access.
- Who can do it
- An owner or admin of the workspace · Open the screen
sso_email_domain_mismatch
- What it means
- The identity provider signed the person in with an e-mail address outside the workspace’s verified domain, or with none, so no session was opened.
- Most frequent cause
- An account whose e-mail is in another domain — a subdomain counts as another —, or a provider that does not release the e-mail.
- What to do
- Sign in with your organisation’s account; an administrator makes the identity provider release an e-mail in the verified domain.
- Who can do it
- An owner or admin of the workspace · Open the screen
sso_email_unverified
- What it means
- The identity provider did not confirm it verified the e-mail address, so no session was opened.
- Most frequent cause
- No email_verified claim; on Microsoft Entra ID, the optional claim xms_edov not added; on Google, an account outside the Workspace domain.
- What to do
- An administrator makes the identity provider confirm the e-mail address (Microsoft Entra ID: add the optional claim xms_edov), then you sign in again.
- Who can do it
- An owner or admin of the workspace
dns_lookup_failed
- What it means
- The domain’s DNS could not be read, so its ownership could not be checked.
- Most frequent cause
- A brief DNS problem, or a domain that does not exist.
- What to do
- Verify the domain again in a minute.
- Who can do it
- An owner or admin of the workspace · Open the screen
record_not_found
- What it means
- The TXT record that proves the domain was not found in its DNS.
- Most frequent cause
- The record not added yet, or not visible yet: DNS changes take a while to spread.
- What to do
- Add the TXT record Settings › Access shows to the domain’s DNS, then verify again a little later.
- Who can do it
- An owner or admin of the workspace · Open the screen
domain_already_verified
- What it means
- Another Azoth workspace has already verified this domain: one domain signs in to one workspace.
- Most frequent cause
- Another workspace — of your organisation or not — proved the domain first.
- What to do
- Use the workspace that holds the domain, or contact support if the domain is yours.
- Who can do it
- An owner or admin of the workspace · Open the screen
domain_changed
- What it means
- The single sign-on domain changed while it was being verified, so nothing was recorded.
- Most frequent cause
- Two people editing single sign-on at the same time.
- What to do
- Reload Settings › Access, then verify the domain again.
- Who can do it
- An owner or admin of the workspace · Open the screen
sso_provider_update_failed
- What it means
- The domain is verified, but registering it with the sign-in service failed.
- Most frequent cause
- A brief problem of the sign-in service.
- What to do
- Verify the domain again in a moment.
- Who can do it
- Azoth, or the operator of this deployment
sso_provider_delete_failed
- What it means
- The sign-in service did not remove the workspace’s SAML identity provider, so Azoth removed nothing either: single sign-on stays as it was.
- Most frequent cause
- A brief problem of the sign-in service.
- What to do
- The sign-in service did not remove the identity provider, so nothing was removed. Try again in a moment.
- Who can do it
- Azoth, or the operator of this deployment
link_expired
- What it means
- This link to connect your chat account has expired.
- Most frequent cause
- The link was opened too late, or used already.
- What to do
- Ask the Azoth app for a new link — /azoth link in Slack, link in Teams — and open it straight away.
- Who can do it
- Any member of the workspace
not_installed
- What it means
- This Slack workspace or Teams bot is not connected to Azoth.
- Most frequent cause
- The app was removed, or was never connected to this deployment.
- What to do
- An owner or admin connects the app in Settings › Integrations and alerts; then ask the app for a new link.
- Who can do it
- An owner or admin of the workspace · Open the screen
chat_identity_mismatch
- What it means
- This chat account is not yours: the e-mail its platform holds for it is not the e-mail of your Azoth account, so nothing was linked.
- Most frequent cause
- A link opened while signed in to another Azoth account, or someone else’s link.
- What to do
- Open the link signed in to the Azoth account that has the chat account’s e-mail: only its owner links it.
- Who can do it
- Any member of the workspace
chat_identity_unverified
- What it means
- The chat platform does not give this account’s e-mail, so Azoth cannot tell whose it is: nothing was linked.
- Most frequent cause
- Slack connected without access to its members’ e-mails, or a Teams guest account.
- What to do
- For Slack, an owner or admin reconnects the app in Settings › Integrations and alerts, then try again.
- Who can do it
- An owner or admin of the workspace · Open the screen
teams_secret_required
- What it means
- A Teams bot is saved with its client secret.
- Most frequent cause
- The client secret left empty.
- What to do
- Paste the bot’s client secret from Microsoft Entra, then save.
- Who can do it
- An owner or admin of the workspace · Open the screen
teams_credentials_invalid
- What it means
- Microsoft refused these credentials, so the bot was not saved.
- Most frequent cause
- A wrong App ID, tenant ID or client secret, or a secret that has expired.
- What to do
- Check the App ID, the tenant ID and the client secret in Microsoft Entra, then save again.
- Who can do it
- An owner or admin of the workspace · Open the screen
teams_app_taken
- What it means
- This Teams bot is already connected to another Azoth workspace.
- Most frequent cause
- The same App ID given to two workspaces.
- What to do
- Give this workspace a bot of its own, or disconnect this one from the other workspace first.
- Who can do it
- An owner or admin of the workspace · Open the screen
connector_not_available
- What it means
- Connecting an assistant (Claude, ChatGPT) is not included in this workspace’s plan.
- Most frequent cause
- The workspace is on a plan without connected assistants.
- What to do
- Move to a plan that includes connected assistants (Settings › Organization), then connect the assistant again.
- Who can do it
- An owner or admin of the workspace · Open the screen
connectors_disabled
- What it means
- The workspace’s owners or admins turned connected assistants off.
- Most frequent cause
- Connected assistants are off in Settings › Access.
- What to do
- Ask an owner or admin to turn connected assistants on in Settings › Access.
- Who can do it
- An owner or admin of the workspace · Open the screen
missing_token
- What it means
- The connector endpoint was called without an access token.
- Most frequent cause
- An assistant not connected to Azoth yet, or a call made by hand.
- What to do
- Connect the assistant to Azoth from its own settings: it signs in and receives its token.
- Who can do it
- The developer of the calling code
invalid_token
- What it means
- The connector endpoint was called with an access token it does not accept: expired, revoked, malformed or issued for another address.
- Most frequent cause
- A connection the person removed, a token past its life, or a deployment whose address changed.
- What to do
- Connect the assistant to Azoth again from its own settings: its access token expired or was revoked.
- Who can do it
- The developer of the calling code
insufficient_scope
- What it means
- The access token does not carry the read access the connector needs.
- Most frequent cause
- A connection granted without read access to the workspace.
- What to do
- Connect the assistant to Azoth again from its own settings, and grant it read access to the workspace.
- Who can do it
- The developer of the calling code
missing_client
- What it means
- The connection request names no client, so the assistant cannot be connected.
- Most frequent cause
- An incomplete link to the consent page.
- What to do
- Start the connection again from the assistant’s settings.
- Who can do it
- The developer of the calling code
unknown_client
- What it means
- The connection request names an assistant Azoth does not know.
- Most frequent cause
- An assistant never registered with this deployment, or whose registration was removed.
- What to do
- Start the connection again from the assistant’s settings, so that it registers again.
- Who can do it
- The developer of the calling code
invalid_redirect_uri
- What it means
- The connection request’s return address is not one the assistant declared, so Azoth will not send it there.
- Most frequent cause
- A link that was modified, or an assistant whose return address changed.
- What to do
- Start the connection again from the assistant’s settings.
- Who can do it
- The developer of the calling code
connector_not_installed
- What it means
- The Inventory Object connector is not installed for this workspace and environment, so nothing was stored.
- Most frequent cause
- The connector was never installed, or was suspended, for the environment the batch names.
- What to do
- Install the connector for this environment (Connectors), or send the batch to an environment where it is installed.
- Who can do it
- An owner or admin of the workspace · Open the screen
connector_revoked
- What it means
- This connector was uninstalled, so it cannot be approved: an uninstall is never turned back on.
- Most frequent cause
- An owner or admin uninstalled it — before the approval, or while a four-eyes request for it was waiting.
- What to do
- Install the connector again: it starts over as a new installation, and waits for approval if this workspace requires one.
- Who can do it
- An owner or admin of the workspace · Open the screen
connector_governance_failed
- What it means
- The connector’s governance check failed during the test ingestion, so the test row was not stored.
- Most frequent cause
- A configuration problem of the connector, or a brief database problem.
- What to do
- Run the test again in a moment. If it keeps failing, contact support: the server’s log names the cause.
- Who can do it
- Azoth, or the operator of this deployment
repository_not_found
- What it means
- No repository of this workspace’s GitHub connection has this id.
- Most frequent cause
- A repository the GitHub App no longer reaches, or an id from another workspace.
- What to do
- Reload the repositories in Settings › Integrations and alerts, then choose one of them.
- Who can do it
- An owner or admin of the workspace · Open the screen