Understand error codes

AIModelActionRequestTimeout

Error message: The prompt prompt-name execution timed out.

Resolution: Make sure that the call to the AI Builder model doesn't exceed 100 seconds.

AsyncResponsePayloadTooLarge

Error message: The output returned from the connector was too large to be handled by the agent. Try reducing its size by utilizing available connector filters or by limiting the number of configured action outputs.

Resolution: One of the agent's real-time connectors is returning a payload that's larger than the agent can handle. For more information about the payload limit, see Copilot Studio web app limits.

ContentError

Error message: This error produces dynamic messages based on the context of the error.

Resolution: This message is a catch-all error for problems that are related to your agent's content. The error message provides more details.

Common problems include:

• A node is missing required properties.
• Invalid YAML was added by using the code editor.
• A Power Fx formula contains an error.

ConsentNotProvidedByUser

Error message: No consent provided for SSO connection.

Resolution: The user who's interacting with the agent must verify the connection by using the agent's single sign-on connection prompt.

DataLossPreventionViolation

Error message: This environment requires users to sign in before they can use the agent. Go to Manage > Security > Authentication and select the option that requires users to sign in.

Resolution

• Your environment's data policies require that users sign in. See Add user authentication with the Sign in system topic.
• One or more connectors that you use in the agent aren't in the same data group. See Copilot Studio connectors.
• One or more connectors that you use in the agent were blocked by the tenant administrator.

EnforcementMessageC2

Error message: This agent is currently unavailable. It has reached its usage limit. Please try again later.

Resolution: This message appears when an agent reaches its message capacity or the pay-as-you-go meter reaches its limit. To resolve the issue, add more prepaid capacity or create a pay-as-you-go billing plan. The agent chat should then resume working within five minutes. For more information, see Overage Enforcement.

FlowActionBadRequest

Error messages:

• The parameter with name {KeyName} on flow {FlowName} ({FlowId}) is declared to be of type {ItemTypeKind}. This type isn't supported when invoking Power Automate. Currently, only Text, Boolean and Numbers are supported.
• The parameter with name {ItemKey} on flow {FlowName} ({FlowId}) is missing in the 'Call Flow' action.
• The parameter with name {KeyName} on flow {FlowName} ({FlowId}) evaluated to type {ResolveType}, expected type {ExpectedType}.
• The flow {FlowName} ({FlowId}) failed to run with response code {ResponseCode}, error code: {FlowErrorCode}.

Resolution: Verify that the base type of any variables that you pass to the flow matches the parameter type.

FlowActionException

Error messages:

• No output was received from flow {FlowName} ({FlowId}), even though output was expected as per the agent definition.
• The output parameter with name {ItemKey} on flow {FlowName} ({FlowId}) is missing from the response data. Refresh the flow, or ensure the flow returns this value.
• The output parameter with name {ItemKey} on flow {FlowName} ({FlowId}) is missing from the output schema. Please refresh the flow.

Resolution: Check the flow for errors.

FlowActionTimedOut

Error message: The flow with id {FlowId} has timed out. Error Code: {FlowErrorCode}

Resolution: Check the flow for errors to understand why the cloud flow took more than 100 seconds to run before it returned to your agent. Try to optimize the query and the data that you return from the back-end system. If some of the cloud flow logic can continue to run after a result is sent to the agent, put these actions after the "Return value(s) to Copilot Studio" step in your cloud flow.

FlowMakerConnectionBlocked

Error message: The flow with name {FlowName} is using a maker connection, which isn't allowed. Error Code: {FlowMakerConnectionBlocked}

Resolution: The administrator prevents using maker credentials in a connection that's invoked from the agent flow. Open the flow in Power Automate and share the cloud flow by using run-only permissions.

GenAISearchandSummarizeRateLimitReached

Error message: The usage limit for search and summarize has been reached. Please try again later.

Resolution: The agent returns this message when it reaches its generative AI limit to search and summarize sources. For more information, see Resolve throttling errors in agents.

GenAIToolPlannerRateLimitReached

Error message: The usage limit for generative orchestration has been reached. Please try again later.

Resolution: The agent returns this message when it reaches its generative orchestration limit. For more information, see Resolve throttling errors in agents.

InfiniteLoopInBotContent

Error message: Action {DialogId}.{TriggerId}.{ActionId} was executed more than {MaxTurnCount} times in a row. This indicates a cycle in execution of the dialog and hence dialog execution will be terminated.

Resolution: Make sure that the topic ends correctly and links to other topics that end correctly, such as the Escalate system topic.

InvalidContent

Error message: A total of {TotalComponents} component(s) exist in the agent, but none are valid.

Resolution: Open the code editor to review problems that affect the content.

LatestPublishedVersionNotFound

Error message: Unable to retrieve the latest published version of the agent.

Resolution: Publish the agent.

OpenAIHate

Error message: The content was filtered due to Responsible AI restrictions.

A Responsible AI check blocks the content for hateful content. Hateful content refers to any content that attacks or uses discriminatory language that refers to a person or identity group based on certain differentiating attributes of these groups.

This restriction includes, but isn't limited to, content about:

• Race, ethnicity, nationality
• Gender identity groups and expression
• Sexual orientation
• Religion
• Personal appearance and body size
• Disability status
• Harassment and bullying

Resolution: To avoid this situation, reinforce responsible AI guidelines together with your agent users. You can also update the agent content moderation policies.

OpenAIJailBreak

Error message: The content was filtered due to Responsible AI restrictions.

A security check blocks the content for a jailbreak attempt. A jailbreak attempt is a user prompt attack that ignores system prompts and tries to alter the intended agent behavior. These attacks include attempts to change system rules, embedding a conversation mockup to confuse the model, role-play, and encoding attacks. For more information, see Prompt Shields in Azure AI Content Safety.

Resolution: To avoid this situation, reinforce responsible AI guidelines together with your agent users. You can also update the agent content moderation policies.

OpenAIndirectAttack

Error message: The content was filtered due to Responsible AI restrictions.

An attack was detected from information that's not directly supplied by the agent author or the user, such as external documents. Attackers attempt to embed instructions in grounded data that's provided by the user in order to maliciously gain control of the system by:

• Manipulating content
• Intrusion
• Unauthorized data exfiltration or data removal from a system
• Blocking system capabilities
• Fraud
• Code execution and infecting other systems

For more information, see Prompt Shields for documents.

Resolution: If you're testing, and you didn't intend the test to be an attack, make sure that your instructions align with what you want the agent to be able to do. Otherwise, reinforce responsible AI guidelines together with your agent users to avoid this situation.

OpenAISelfHarm

Error message: The content was filtered due to Responsible AI restrictions.

A Responsible AI check blocked content that's related to self-harm. Self-harm describes language that's related to physical actions that are intended to purposely hurt, injure, damage one's body, or kill oneself.

This restriction includes, but isn't limited to, content about:

• Eating disorder
• Bullying and intimidation

Resolution: To avoid this situation, reinforce responsible AI guidelines together with your agent users. You can also update the agent content moderation policies.

OpenAISexual

Error message: The content was filtered due to Responsible AI restrictions.

A Responsible AI check blocked content for sexual content. Sexual content describes language that's related to anatomical organs and genitals, romantic relationships, sexual acts, and acts that are portrayed in erotic or affectionate terms, including those portrayed as an assault or a forced sexual violent act against one's will.

This restriction includes, but isn't limited to, content about:

• Vulgar content
• Prostitution
• Nudity and pornography
• Abuse
• Child exploitation, child abuse, child grooming

Resolution: To avoid this situation, reinforce responsible AI guidelines together with your agent users. You can also update the agent content moderation policies.

OpenAIRateLimitReached

Error message: An error has occurred.

Resolution: Your agent reached the maximum number of generative answers responses. Review your message capacity, and review the information in Resolve throttling errors in agents.

OpenAIViolence

Error message: The content was filtered due to Responsible AI restrictions.

A Responsible AI check blocked content for violent content. Violent content describes language that's related to physical actions intended to hurt, injure, damage, or kill someone or something. It also describes weapons, guns, and related entities.

This restriction includes, but isn't limited to, content about:

• Weapons
• Bullying and intimidation
• Terrorist and violent extremism
• Stalking

Resolution: To avoid this situation, reinforce responsible AI guidelines together with your agent users. You can also update the agent content moderation policies.

OutgoingMessageSizeTooBig

Error message: Outgoing message size too big.

Resolution: Depending on the channel that you use to transfer files, such as Direct Line or Facebook, you might receive the following error message: "The request content length exceeded limit of 262,144 bytes." These limits are imposed by the channel, not by Copilot Studio.

In this scenario, consider a few options. One option is to provide a link to the resource as an internet attachment. Another option is to review your nodes to make sure that none of them use a variable that contains a large volume of text, such as a JSON.stringify() static method. If you use this method or a variable that contains a large volume of text, modify the node to pass only the portion of text that's necessary. For example, if you use an Adaptive Card to pass data to another topic, update the variable to pass only the necessary property.

For more information, see Maximum channel data message size limits when using Copilot Studio in Omnichannel.

RedirectToDisabledDialog

Error message: The Dialog with Id {DialogId} is disabled in the definition. Please enable the Dialog before using it.

Resolution: Re-enable the topic or remove the redirect node.

RedirectToNonExistentDialog

Error message: The Dialog with Id {DialogId} wasn't found in the definition. Please check that the Dialog is present and that the Id is correct.

Resolution: Create a new topic to redirect to, or remove the redirect node.

SystemError

Error message: This error doesn't produce an error message.

Resolution: Contact customer support.

TooMuchDataToHandle

Error message: The request is resulting in too much data to handle, please evaluate the amount of data being returned by your actions.

Resolution: This error indicates that the request that was sent to OpenAI exceeds the maximum request size allowed. The request includes the user input, output from previous actions, tools that were called, and conversation history. Review the tools that you use. If it's possible, scope down their output to only the necessary fields. For more information, see Create a Power Automate flow and Call a Power Automate flow as an action.