Map of headers that were added by the action.
Overall outcome of the action: "success" or "error".
Error details. Only present when status is "error".
Error code (e.g., "ERR\_NGROK\_3807").
Error message describing the failure.
Information about the client request.
HTTP method from the client request.
Request path from the client.
Client request headers (API key/auth headers trimmed to avoid leaking).
User-Agent header from the client.
Client-requested model field.
Client-requested models field.
Trimmed client-supplied api key to avoid leaking while still being able to correlate it with the full key if you have it.
Models the client requested but were excluded.
Model the client requested but was excluded.
Explanation for why it was rejected.
Input token count calculated by ngrok (used for billing).
Output token count calculated by ngrok (used for billing).
Time added by gateway logic in milliseconds. Excludes time the gateway spend waiting on the upstream provider(s) to respond.
Details about the model selection process.
The resolved strategy expression.
Error if strategy evaluation failed.
Models returned by the strategy expression.
List after .disabled, only\_allow\_configured\_\* filters.
List after filtering by client-requested models.
Final list of models to try.
Models attempted in chronological order.
Model identifier.
Provider Provider identifier for the model tried.
Author identifier for the model tried (may be empty).
Details about API key selection for this model.
Strategy used for API key selection.
Error if strategy failed.
List of trimmed keys to try.
List of requests made for this model.
Request outcome: "success" or "error".
Error message if request failed.
Input token count reported by the upstream provider.
Output token count reported by the upstream provider.
Details of the request made by the gateway
Request URL.
Partially redacted API key for safe display and correlation with the full key.
Response details.
HTTP status code.
Response headers.
Only populated on failure, truncated to 1KB.
Upstream latency measurements.
Milliseconds until first byte of response received.
Milliseconds until request fully completed.
Details of the successful model. Only present when status is "success".
Index into the models\_tried array for the successful model.
Index into requests\_made array for the successful request.
Model that successfully handled the request (e.g., "gpt-4").
Provider identifier for the successful model.
Author identifier for the successful model. May be empty.
Whether there were Basic Auth credentials presented in the Authorization header.
The username of the credentials presented.
Whether the presented basic auth credentials were authorized for this request.
Indicates whether the body was already compressed before the action was
applied. Returns true if no further compression was
performed.
The compression algorithm selected and applied by the action, based on the client's request and action configuration.
Message for an error that occurred during the invocation of an action.
A list of HTTP responses for each request attempt.
The IP address of the host returning the response.
A map of HTTP response headers.
The HTTP Status code for the response attempt.
Time it took to receive the response.
The HTTP request.
The HTTP method of the request.
A map of HTTP request headers.
The request url.
The request body.
The last attempted HTTP response. Unlike `actions.ngrok.attempts[i]` this variable also contains the response body.
The IP address of the host returning the response.
A map of HTTP response headers.
The HTTP Status code for the response.
Time it took to receive the response.
The response body
The list of JSON Web Tokens (JWTs) processed by the action.
The header portion of the JWT, parsed into a key-value map containing metadata about the token.
The part of the HTTP request where the JWT was located (for example,{" "}
header, body).
The specific header or body field that contained the JWT (for example,{" "}
"Authorization").
The payload portion of the JWT, parsed into a key-value map representing the claims.
The signature portion of the JWT, encoded in base64, used for verifying token integrity.
Indicates whether the token passed verification. Returns{" "}
true if verified successfully.
A machine-readable code describing an error that occurred during the action's execution.
A human-readable message providing details about an error that occurred during the action's execution.
A key-value map containing metadata that was logged during the action. Each key represents a metadata attribute, and the value provides its corresponding details.
Code for an error that occurred during the invocation of an action.
Message for an error that occurred during the invocation of an action.
Unique identifier for the ngrok Identity entity
Email address of the authorized user from the provider.
Name for the authorized user from the provider.
Identifier for the authorized user from the provider.
The current Identity session identifier for this request.
The access token from the provider.
The refresh token from the provider.
Timestamp when the current session will expire.
Returns true when the session timed out.
Returns true when the current session reached the max duration.
Returns true when ngrok updates the user information on the identity.
Code for an error that occurred during the invocation of an action.
Message for an error that occurred during the invocation of an action.
Unique identifier for the ngrok Identity entity
Email address of the authorized user from the provider.
Name for the authorized user from the provider.
Identifier for the authorized user from the provider.
The current Identity session identifier for this request.
The identity token from the provider for the current user.
The access token from the provider.
The refresh token from the provider.
Timestamp when the current session will expire.
Returns true when the session timed out.
Returns true when the current session reached the max duration.
Returns true when ngrok updates the user information on the identity.
The key used for bucketing requests. This is the key used to group and track requests in the rate-limiting process, ensuring that the same bucket is subject to the rate limit across multiple requests.
Indicates whether the request was limited by the rate limit. If `true`, the request was rate-limited based on the configured limits for the specified bucket.
A machine-readable code describing an error that occurred during the action's execution.
A human-readable message providing details about an error that occurred during the action's execution.
A list of elements that were matched during redirection. These represent the request components (for example, path or query parameters) that triggered the action and resulted in the redirect.
The URL to which the traffic was redirected. This is the destination URL returned as part of the redirect response after the action was executed.
A machine-readable code describing an error that occurred during the action's execution.
A human-readable message providing details about an error that occurred during the action's execution.
A list of headers that were successfully removed by the action.
The action taken for this request.
The CIDR block that matched the incoming client's IP address. If no match was found, this value will be empty.
A machine-readable code describing an error that occurred during the action's execution.
A human-readable message providing details about an error that occurred during the action's execution.
List of elements that matched the URL before the rewrite action was applied. These can be specific parts of the URL, such as the domain, path, or query parameters, that were matched based on the action configuration.
The final URL after the rewrite action has been applied. This is the new URL to which the original request is redirected after the specified modifications have been made.
A machine-readable code describing an error that occurred during the action's execution.
A human-readable message providing details about an error that occurred during the action's execution.
Indicates whether or not the request was successfully verified.
Code for an error that occurred during the invocation of an action.
Message for an error that occurred during the invocation of an action.