AI Guard | Node.js SDK

AI Guard

constructor(token: string, config: PangeaConfig): AIGuardService

Creates a new AIGuardService with the given Pangea API token and configuration.

required parameters

token

string

Pangea API token.

configPangeaConfig

Configuration.

Response Object

AIGuardService

const config = new PangeaConfig({ domain: "pangea_domain" });
const aiGuard = new AIGuardService("pangea_token", config);

Text guard (Beta)

guardText(request: TextGuardRequest): Promise<PangeaResponse<TextGuardResult>>

Undocumented.

required parameters

requestTextGuardRequest

Request parameters.

Response Object

TextGuardResult

Promise<PangeaResponse<TextGuardResult>>

const response = await aiGuard.guardText({
  text: "foobar",
});

Interface IntelResults

IntelResults

required parameters

category

Array<string>

The categories that apply to this indicator as determined by the provider.

score

number

The score, given by the Pangea service, for the indicator.

verdict

malicious | suspicious | unknown | benign

The verdict for the indicator.

Interface RedactRecognizerResult

RedactRecognizerResult

required parameters

end

number

The ending index of a snippet.

field_type

string

The entity name.

redacted

boolean

Indicates if this rule was used to anonymize a text snippet.

score

number

The certainty score that the entity matches this specific snippet.

start

number

The starting index of a snippet.

text

string

The text snippet that matched.

Interface RedactReport

RedactReport

required parameters

count

number

recognizer_resultsArray<RedactRecognizerResult>

Interface TextGuardArtifact

TextGuardArtifact

required parameters

defanged

boolean

end

number

start

number

type

string

value

string

optional parameters

verdict

string

The verdict, given by the Pangea service, for the indicator.

Interface TextGuardFindings

TextGuardFindings

required parameters

artifact_count

number

malicious_count

number

security_issuesTextGuardSecurityIssues

Interface TextGuardReport

TextGuardReport

required parameters

redactRedactReport

optional parameters

domain_intelIntelResults

ip_intelIntelResults

url_intelIntelResults

user_intelBreachedData

Interface TextGuardRequest

TextGuardRequest

required parameters

text

string

optional parameters

debug

boolean

recipe

[object Object] | pangea_prompt_guard

Interface TextGuardResult

TextGuardResult

required parameters

findingsTextGuardFindings

redacted_prompt

string

optional parameters

artifactsArray<TextGuardArtifact>

reportTextGuardReport

Interface TextGuardSecurityIssues

TextGuardSecurityIssues

required parameters

compromised_email_addresses

number

malicious_domain_count

number

malicious_ip_count

number

malicious_url_count

number

redact_rule_match_count

number