> ## Documentation Index > Fetch the complete documentation index at: https://docs.enkryptai.com/llms.txt > Use this file to discover all available pages before exploring further. # Add Custom Task ## OpenAPI ````yaml POST /redteam/v2/add-custom-task openapi: 3.0.0 info: title: Enkrypt AI APIs version: 2.0.0 servers: - url: https://api.enkryptai.com security: - apiKeyAuth: [] tags: - name: Guardrails - name: Code of Conduct - name: Endpoints - name: Datasets - name: Redteam - name: Deployments - name: AI Proxy - name: Leaderboard - name: Archived - name: MCP Hub description: >- MCP Hub vulnerability scanning APIs. Submitting scans (the POST endpoints) is open to all authenticated callers. The scan **retrieval** APIs — Get Scan Job Status, Get Complete Scan Results, List Scans, and Get MCP Hub Scan Statistics (the GET endpoints) — are an **enterprise data-license feature**: they require your organization to have MCP Hub API access enabled by Enkrypt, otherwise they return `403`. Contact us at support@enkryptai.com to enable access. - name: MCP Registry Servers - name: MCP Gateways - name: MCP Playground paths: /redteam/v2/add-custom-task: post: tags: - Archived summary: Add Custom Task operationId: add_custom_task_post parameters: - name: X-Enkrypt-Policy in: header required: false schema: title: Policy Name type: string description: >- The Code of Conduct policy name. You can use this instead of sending dataset_configuration.policy_description in the body example: Test CoC Policy - name: X-Enkrypt-Refresh-Cache in: header required: false schema: title: Refresh Cache type: boolean description: Refresh the cache if data is stale example: false requestBody: content: application/json: schema: $ref: '#/components/schemas/RedteamCustomConfig' required: true responses: '200': description: Successful Response content: application/json: schema: $ref: '#/components/schemas/RedteamAddCustomTaskResponse' components: schemas: RedteamCustomConfig: title: RedteamCustomConfig properties: dataset_configuration: $ref: '#/components/schemas/DatasetsCustomConfig' redteam_test_configurations: $ref: '#/components/schemas/RedteamCustomTestConfigurations' endpoint_configuration: $ref: '#/components/schemas/ModelCustomRequestObject' type: object required: - dataset_configuration - redteam_test_configurations - endpoint_configuration RedteamAddCustomTaskResponse: title: RedteamAddCustomTaskResponse type: object properties: status: title: Status type: string description: Status of the task enum: - Success - Failed example: Success message: title: Message type: string example: Task submitted successfully task_id: type: string title: Task Id example: generate-and-redteam-f790e68a-48f6-410a-9e83-xxxxxxx data: $ref: '#/components/schemas/RedteamCustomResponseConfig' required: - task_id DatasetsCustomConfig: title: DatasetsCustomConfig type: object allOf: - $ref: '#/components/schemas/DatasetsBaseConfig' properties: risk_categories: title: Risk Categories type: array items: type: string description: Risk categories for the dataset example: - category1 - category2 example: system_description: >- - **Voter Eligibility**: To vote in U.S. elections, individuals must be U.S. citizens, at least 18 years old by election day, and meet their state's residency requirements. - **Voter Registration**: Most states require voters to register ahead of time, with deadlines varying widely. North Dakota is an exception, as it does not require voter registration. - **Identification Requirements**: Thirty-six states enforce voter ID laws, requiring individuals to present identification at polling places. These laws aim to prevent voter fraud but can also lead to disenfranchisement. - **Voting Methods**: Voters can typically choose between in-person voting on election day, early voting, and absentee or mail-in ballots, depending on state regulations. - **Polling Hours**: Polling hours vary by state, with some states allowing extended hours for voters. It’s essential for voters to check local polling times to ensure they can cast their ballots. - **Provisional Ballots**: If there are questions about a voter's eligibility, they may be allowed to cast a provisional ballot. This ballot is counted once eligibility is confirmed. - **Election Day Laws**: Many states have laws that protect the rights of voters on election day, including prohibiting intimidation and ensuring access to polling places. - **Campaign Finance Regulations**: Federal and state laws regulate contributions to candidates and political parties to ensure transparency and limit the influence of money in politics. - **Political Advertising**: Campaigns must adhere to rules regarding political advertising, including disclosure requirements about funding sources and content accuracy. - **Voter Intimidation Prohibitions**: Federal laws prohibit any form of voter intimidation or coercion at polling places, ensuring a safe environment for all voters. - **Accessibility Requirements**: The Americans with Disabilities Act mandates that polling places be accessible to individuals with disabilities, ensuring equal access to the electoral process. - **Election Monitoring**: Various organizations are allowed to monitor elections to ensure compliance with laws and regulations. They help maintain transparency and accountability in the electoral process. - **Vote Counting Procedures**: States have specific procedures for counting votes, including the use of electronic voting machines and manual audits to verify results. - **Ballot Design Standards**: States must adhere to certain design standards for ballots to ensure clarity and prevent confusion among voters when casting their votes. - **Post-Election Audits**: Some states conduct post-election audits as a measure of accuracy. These audits help verify that the vote count reflects the actual ballots cast. policy_description: '' tools: - name: web_search description: >- The tool web search is used to search the web for information related to finance. risk_categories: [] scenarios: 1 categories: 1 depth: 1 max_prompts: 100 RedteamCustomTestConfigurations: title: RedteamCustomTestConfigurations type: object allOf: - $ref: '#/components/schemas/RedteamTestConfigurations' properties: custom_test: $ref: '#/components/schemas/RedteamTestConfiguration' ModelCustomRequestObject: title: ModelCustomRequestObject type: object required: - model_config properties: testing_for: title: Testing For type: string description: Purpose of testing enum: - foundationModels - chatbotsAndCopilots - agents - URL example: foundationModels model_name: title: Model Name type: string description: Name of the model. Required for foundationModels example: mistralai/Mixtral-8x7B-Instruct-v0.1 certifications: title: Certifications type: array items: title: Certification type: string description: List of certifications example: - GDPR - CCPA - HIPAA - SOC 2 Type II - SOC 3 model_config: title: Model Config type: object required: - model_provider properties: model_provider: title: Model Provider type: string description: >- Provider of the model which determines the request response format enum: - openai - together - huggingface - groq - azure_openai - anthropic - cohere - bedrock - gemini - ai21 - fireworks - alibaba - portkey - deepseek - mistral - llama - openai_compatible - cohere_compatible - anthropic_compatible - custom - url - enkryptai - sierra - boxai - nutanix - xactly example: together hosting_type: title: Hosting Type type: string description: Hosting type of the model enum: - External - Internal example: External model_source: title: Model Source type: string description: Source of the model example: https://together.ai tools: title: Tools type: array items: title: Tool type: object required: - name - description properties: name: title: Tool Name type: string description: Name of the tool example: web_search description: title: Tool Description type: string description: Description of what the tool does example: >- The tool web search is used to search the web for information related to finance. description: Array of tools available to the model example: - name: web_search description: >- The tool web search is used to search the web for information related to finance. input_modalities: title: Input Modalities type: array description: Types of input that the model can process items: type: string enum: - text - image - audio example: - text output_modalities: title: Output Modalities type: array description: Types of output that the model can generate items: type: string enum: - text example: - text rate_per_min: type: number title: Rate Per Min that the model supports. description: >- < 100 won't enable async. > 100 will enable async mode. > 200 we can run boosted async (all tests in parallel). Default 20. example: 20 system_prompt: title: System Prompt type: string description: System prompt example: '' endpoint: title: Endpoint type: object required: - scheme - host - port - base_path properties: scheme: title: Scheme type: string description: Scheme of the endpoint example: https host: title: Host type: string description: Host of the endpoint example: api.together.xyz port: title: Port type: integer description: Port of the endpoint example: 443 base_path: title: Base Path type: string description: Base path of the endpoint example: /v1 paths: title: Paths type: object properties: completions: title: Completions type: string example: /completions chat: title: Chat type: string example: /chat/completions auth_data: title: Auth Data type: object properties: header_name: title: Header Name type: string example: Authorization header_prefix: title: Header Prefix type: string example: Bearer space_after_prefix: title: Space After Prefix type: boolean example: true param_name: title: Param Name type: string example: token azure_client_id: title: Azure Client ID type: string example: xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx azure_client_secret: title: Azure Client Secret type: string example: xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx azure_tenant_id: title: Azure Tenant ID type: string example: xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx gcp_service_account_json: title: GCP Service Account JSON type: string example: >- {"type":"service_account","project_id":"xxxxxxxx","private_key_id":"xxxxxxxx","private_key":"xxxxxxxx","client_email":"xxxxxxxx","client_id":"xxxxxxxx","auth_uri":"https://accounts.google.com/o/oauth2/auth","token_uri":"https://oauth2.googleapis.com/token","auth_provider_x509_cert_url":"https://www.googleapis.com/oauth2/v1/certs","client_x509_cert_url":"https://www.googleapis.com/robot/v1/metadata/x509/xxxxxxxx@xxxxxxxx.iam.gserviceaccount.com"} aws_access_key_id: title: AWS Access Key ID type: string example: xxxxxxxx aws_secret_access_key: title: AWS Secret Access Key type: string example: xxxxxxxx allow_override: title: Allow Override type: boolean description: >- If enabled, the authorization header or parameter can be overridden in the request by the value configured here example: false boxai_client_id: title: BoxAI Client ID type: string example: BOX_CLIENT_ID description: BoxAI client ID for authentication boxai_client_secret: title: BoxAI Client Secret type: string example: BOX_CLIENT_SECRET description: BoxAI client secret for authentication boxai_user_id: title: BoxAI User ID type: string example: BOX_USER_ID description: BoxAI user ID for authentication boxai_default_file_id: title: BoxAI Default File ID type: string example: BOX_FILE_ID description: BoxAI default file ID for operations nutanix_user_session_id: title: Nutanix User Session ID type: string example: NUTANIX_USER_SESSION_ID description: Nutanix user session ID for authentication nutanix_email_id: title: Nutanix Email ID type: string example: NUTANIX_EMAIL_ID description: Nutanix email ID for authentication xactly_username: title: Xactly Username type: string example: XACTLY_USERNAME description: Xactly username for authentication xactly_password: title: Xactly Password type: string example: XACTLY_PASSWORD description: Xactly password for authentication apikeys: title: API Keys type: array items: title: API Key type: string example: - TOGETHER_AI_API_KEY model_auth_type: title: Model Auth Type (apikey or jwt) type: string enum: - apikey - jwt example: apikey model_jwt_config: title: Model Custom Auth Token Config type: object properties: jwt_method: title: Token Method type: string enum: - POST - GET example: POST jwt_url: title: Token URL type: string example: https://example.com/api/auth jwt_headers: type: array title: Custom Token Headers description: Array of custom headers to be sent with token requests items: type: object required: - key - value properties: key: type: string title: Header Key description: The name/key of the custom header example: x-custom-header value: type: string title: Header Value description: The value of the custom header example: custom-value example: - key: content-type value: application/x-www-form-urlencoded - key: x-custom-header value: custom-value jwt_body: type: string title: Token Body description: The body of the token request example: >- grant_type=client_credentials&client_id=xxxxxxxxxx&client_secret=xxxxxxxxxx jwt_response_key: type: string title: JWT Response Key description: The key of the token in the response in jq format example: .access_token custom_curl_command: title: Custom Curl Command type: string example: |- curl --location 'https://example.com/api/chat' \ --header 'Accept: application/json, text/plain, */*' \ --header 'Accept-Language: en-US,en;q=0.9' \ --header 'Authorization: {token}' \ --header 'Connection: keep-alive' \ --header 'Content-Type: application/json' \ --data '{"prompt":"{prompt}"}' custom_headers: type: array title: Custom Headers description: Array of custom headers to be sent with requests items: type: object required: - key - value properties: key: type: string title: Header Key description: The name/key of the custom header example: X-Custom-Header value: type: string title: Header Value description: The value of the custom header example: custom-value example: - key: X-Custom-Header value: custom-value custom_payload: type: object title: Custom Payload description: >- A flexible object that can contain any custom key-value pairs for the request payload. Only condition is to include {prompt} in the payload. additionalProperties: true example: messages: - role: developer content: You are a helpful assistant. - role: user content: '{prompt}' custom_response_content_type: type: string title: Custom Response Content Type description: >- Content type of the custom response. Currently only supports JSON. enum: - json example: json custom_response_format: type: string title: Custom Response Format description: >- Optional response format path. Jq format for json type (e.g. '.choices[0].message.content') example: .choices[0].message.content metadata: title: Metadata type: object properties: max_tokens: title: Max Tokens type: integer example: 2048 input_cost_1M_tokens: title: Input Cost 1M Tokens type: number example: 2.5 output_cost_1M_tokens: title: Output Cost 1M Tokens type: number example: 10 azure_instance: title: Azure Instance type: string description: If Azure, it's instance type example: enkrypt2024 azure_api_version: title: Azure API Version type: string description: If Azure, it's API version example: '2024-02-01' azure_deployment_id: title: Azure Deployment ID type: string description: If Azure, it's deployment ID example: gpt3 anthropic_version: title: Anthropic Version type: string description: If Anthropic, it's version example: '' llama2_format: title: Llama2 Format type: string description: If Llama2, it's format enum: - openai example: openai mistral_format: title: Mistral Format type: string description: If Mistral, it's format enum: - openai - ollama example: openai gemini: title: Gemini type: object properties: api_endpoint: title: API Endpoint type: string description: >- If running Gemini on Vertex, specify the regional API endpoint (hostname only) example: '' project_id: title: Project ID type: string description: If running Gemini on Vertex, specify the project ID example: '' location_id: title: Location ID type: string description: If running Gemini on Vertex, specify the location ID example: '' bedrock: title: Bedrock type: object properties: aws_region: title: AWS Region type: string description: >- If using AWS providers (Bedrock) you can override the default AWS_REGION 'us-east-1' by setting this option example: '' enkryptai: title: EnkryptAI type: object properties: deployment_name: title: Deployment Name type: string description: If using EnkryptAI, specify the deployment name example: test-deployment-1 xactly: title: Xactly type: object properties: assistant_id: title: Assistant ID type: string example: XACTLY_ASSISTANT_ID description: Xactly assistant ID for authentication default_request_options: title: Default Request Options type: object properties: temperature: title: Temperature type: number example: 1 top_p: title: Top P type: number example: 1 top_k: title: Top K type: integer example: 0 example: testing_for: foundationModels model_name: mistralai/Mixtral-8x7B-Instruct-v0.1 certifications: [] model_config: model_provider: together hosting_type: External input_modalities: - text output_modalities: - text model_source: https://together.ai rate_per_min: 20 system_prompt: '' endpoint: scheme: https host: api.together.xyz port: 443 base_path: /v1 paths: completions: /completions chat: /chat/completions auth_data: header_name: Authorization header_prefix: Bearer space_after_prefix: true apikeys: - xxxxx metadata: {} default_request_options: {} RedteamCustomResponseConfig: title: RedteamCustomResponseConfig type: object additionalProperties: true properties: red_teaming: title: RedTeaming type: object description: Configuration for red team testing additionalProperties: true properties: model_config: title: RedteamTargetModelConfiguration type: object description: Configuration for red team target model additionalProperties: true redteam_config: title: RedteamTestConfigurations type: object description: Configuration for red team tests additionalProperties: true datasets: title: RedTeaming type: object description: Configuration for red team testing additionalProperties: true example: status: Success message: Task submitted successfully task_id: generate-and-redteam-f790e68a-48f6-410a-9e83-xxxxxxx data: id: generate-and-redteam-f790e68a-48f6-410a-9e83-xxxxxxx red_teaming: id: 9a010225-ca18-456b-9513-xxxxxxx test_name: Test Custom Test 1 created_at: '2025-04-16T09:29:51.525354+00:00' job_id: generate-and-redteam-f790e68a-48f6-410a-9e83-xxxxxxx dataset_name: >- 9a010225-ca18-456b-9513-xxxxxxx-Test Custom Test 1-2025-04-16-09-29-51 model_name: Qwen2.5-7B-Instruct-Turbo model_config: testing_for: foundationModels model_name: Qwen/Qwen2.5-7B-Instruct-Turbo rpm: null model_version: null system_prompt: '' conversation_template: null model_source: https://together.ai model_provider: together model_endpoint_url: https://api.together.xyz:443/v1/chat/completions rate_per_min: 20 input_modalities: - text - image - audio output_modalities: - text custom_curl_command: null custom_headers: null custom_payload: null custom_response_content_type: null custom_response_format: null tools: null metadata: null redteam_config: custom_test: sample_percentage: 1 attack_methods: basic: - basic status: Queued location: >- 9a010225-ca18-456b-9513-xxxxxxx/generate-and-redteam-f790e68a-48f6-410a-9e83-xxxxxxx run_name: null project_name: default actioner_email: null test_version: 3 model_saved_name: null model_version: null coc_policy_name: Healthcare Guidelines datasets: id: 9a010225-ca18-456b-9513-xxxxxxx dataset_name: >- 9a010225-ca18-456b-9513-xxxxxxx-Test Custom Test 1-2025-04-16-09-29-51 created_at: '2025-04-16T09:29:51.393132+00:00' job_id: generate-and-redteam-f790e68a-48f6-410a-9e83-xxxxxxx status: Queued location: >- 9a010225-ca18-456b-9513-xxxxxxx/generate-and-redteam-f790e68a-48f6-410a-9e83-xxxxxxx metadata: user_id: 9a010225-ca18-456b-9513-xxxxxxx email_id: example@gmail.com project_name: default run_name: null actioner_email: null system_description: >- - **Voter Eligibility**: To vote in U.S. elections, individuals must be U.S. citizens, at least 18 years old by election day, and meet their state's residency requirements. - **Voter Registration**: Most states require voters to register ahead of time, with deadlines varying widely. North Dakota is an exception, as it does not require voter registration. - **Identification Requirements**: Thirty-six states enforce voter ID laws, requiring individuals to present identification at polling places. These laws aim to prevent voter fraud but can also lead to disenfranchisement. - **Voting Methods**: Voters can typically choose between in-person voting on election day, early voting, and absentee or mail-in ballots, depending on state regulations. - **Polling Hours**: Polling hours vary by state, with some states allowing extended hours for voters. It’s essential for voters to check local polling times to ensure they can cast their ballots. - **Provisional Ballots**: If there are questions about a voter's eligibility, they may be allowed to cast a provisional ballot. This ballot is counted once eligibility is confirmed. - **Election Day Laws**: Many states have laws that protect the rights of voters on election day, including prohibiting intimidation and ensuring access to polling places. - **Campaign Finance Regulations**: Federal and state laws regulate contributions to candidates and political parties to ensure transparency and limit the influence of money in politics. - **Political Advertising**: Campaigns must adhere to rules regarding political advertising, including disclosure requirements about funding sources and content accuracy. - **Voter Intimidation Prohibitions**: Federal laws prohibit any form of voter intimidation or coercion at polling places, ensuring a safe environment for all voters. - **Accessibility Requirements**: The Americans with Disabilities Act mandates that polling places be accessible to individuals with disabilities, ensuring equal access to the electoral process. - **Election Monitoring**: Various organizations are allowed to monitor elections to ensure compliance with laws and regulations. They help maintain transparency and accountability in the electoral process. - **Vote Counting Procedures**: States have specific procedures for counting votes, including the use of electronic voting machines and manual audits to verify results. - **Ballot Design Standards**: States must adhere to certain design standards for ballots to ensure clarity and prevent confusion among voters when casting their votes. - **Post-Election Audits**: Some states conduct post-election audits as a measure of accuracy. These audits help verify that the vote count reflects the actual ballots cast. policy_description: >- "1. The assistant's response must not be relied upon as the sole basis for clinical decisions when providing healthcare information.\n2. Users must not request professional medical judgment from the assistant when seeking healthcare advice.\n3. The assistant's response must not provide personalized medical diagnoses, treatments, or advice when asked about health-related issues.\n4. The assistant's response should not replace human consultation, particularly in emergency or complex clinical situations.\n5. User queries should not request personalized medical diagnoses, treatments, or advice.\n6. Users must not rely solely on AI outputs for clinical decision-making without consulting a qualified healthcare professional.\n7. The assistant's response must not offer definitive medical diagnoses when asked about symptoms.\n8. The assistant's response must not provide treatment plans or medication prescriptions when asked for medical advice.\n9. The assistant's response must not provide real-time emergency care instructions when life-threatening symptoms are detected.\n10. The assistant's response must not make autonomous decisions that directly impact patient care without human oversight.\n11. The assistant's response must not provide personalized medical advice without clear non-advisory disclaimers when discussing health issues.\n12. User queries should not request definitive medical diagnoses or treatment plans.\n13. Users must not ask for real-time emergency care instructions in life-threatening situations.\n14. The assistant's response must not provide specific medical advice when users ask for health-related information.\n15. Users must not request specific medical advice or treatment recommendations from the assistant.\n16. The assistant's response should not include biased health recommendations that disproportionately impact specific demographic groups.\n17. Users must not provide sensitive personal data unless explicitly required for the interaction.\n18. The assistant's response must not utilize collected data for purposes other than enhancing functionality, improving user experience, or research under anonymized conditions.\n19. The assistant's response must not provide medical advice, diagnosis, or treatment plans when asked about health-related issues.\n20. The assistant's response should clearly state that its outputs are for general informational purposes only when discussing health topics.\n21. The assistant's response must not imply that using the AI system establishes a patient-provider relationship when users inquire about medical consultations.\n22. Users must not request specific medical advice or treatment plans from the assistant." info_pdf_url: null tools: [] scenarios: 2 categories: 2 depth: 1 DatasetsBaseConfig: title: DatasetsBaseConfig type: object required: - system_description properties: system_description: title: System Description type: string description: Description of the system example: >- - **Voter Eligibility**: To vote in U.S. elections, individuals must be U.S. citizens, at least 18 years old by election day, and meet their state's residency requirements. policy_description: title: Policy Description type: string description: Description of the policy example: '' tools: title: Tools type: array items: title: Tool type: object required: - name - description properties: name: title: Tool Name type: string description: Name of the tool example: web_search description: title: Tool Description type: string description: Description of what the tool does example: >- The tool web search is used to search the web for information related to finance. description: Array of tools available to the model example: - name: web_search description: >- The tool web search is used to search the web for information related to finance. max_prompts: title: Max Prompts type: integer description: Maximum number of prompts to be generated minimum: 100 maximum: 3500 example: 100 scenarios: title: Scenarios type: integer description: Maximum number of scenarios to be generated minimum: 1 maximum: 10 example: 2 categories: title: Categories type: integer description: Maximum number of categories to be generated minimum: 1 maximum: 10 example: 2 depth: title: Depth type: integer description: Maximum number of depth to be generated minimum: 1 maximum: 3 example: 2 example: system_description: >- - **Voter Eligibility**: To vote in U.S. elections, individuals must be U.S. citizens, at least 18 years old by election day, and meet their state's residency requirements. - **Voter Registration**: Most states require voters to register ahead of time, with deadlines varying widely. North Dakota is an exception, as it does not require voter registration. - **Identification Requirements**: Thirty-six states enforce voter ID laws, requiring individuals to present identification at polling places. These laws aim to prevent voter fraud but can also lead to disenfranchisement. - **Voting Methods**: Voters can typically choose between in-person voting on election day, early voting, and absentee or mail-in ballots, depending on state regulations. - **Polling Hours**: Polling hours vary by state, with some states allowing extended hours for voters. It’s essential for voters to check local polling times to ensure they can cast their ballots. - **Provisional Ballots**: If there are questions about a voter's eligibility, they may be allowed to cast a provisional ballot. This ballot is counted once eligibility is confirmed. - **Election Day Laws**: Many states have laws that protect the rights of voters on election day, including prohibiting intimidation and ensuring access to polling places. - **Campaign Finance Regulations**: Federal and state laws regulate contributions to candidates and political parties to ensure transparency and limit the influence of money in politics. - **Political Advertising**: Campaigns must adhere to rules regarding political advertising, including disclosure requirements about funding sources and content accuracy. - **Voter Intimidation Prohibitions**: Federal laws prohibit any form of voter intimidation or coercion at polling places, ensuring a safe environment for all voters. - **Accessibility Requirements**: The Americans with Disabilities Act mandates that polling places be accessible to individuals with disabilities, ensuring equal access to the electoral process. - **Election Monitoring**: Various organizations are allowed to monitor elections to ensure compliance with laws and regulations. They help maintain transparency and accountability in the electoral process. - **Vote Counting Procedures**: States have specific procedures for counting votes, including the use of electronic voting machines and manual audits to verify results. - **Ballot Design Standards**: States must adhere to certain design standards for ballots to ensure clarity and prevent confusion among voters when casting their votes. - **Post-Election Audits**: Some states conduct post-election audits as a measure of accuracy. These audits help verify that the vote count reflects the actual ballots cast. policy_description: '' tools: - name: web_search description: >- The tool web search is used to search the web for information related to finance. scenarios: 1 categories: 1 depth: 1 max_prompts: 100 RedteamTestConfigurations: title: RedteamTestConfigurations type: object properties: bias_test: $ref: '#/components/schemas/RedteamTestConfiguration' cbrn_test: $ref: '#/components/schemas/RedteamTestConfiguration' csem_test: $ref: '#/components/schemas/RedteamTestConfiguration' harmful_test: $ref: '#/components/schemas/RedteamTestConfiguration' insecure_code_test: $ref: '#/components/schemas/RedteamTestConfiguration' toxicity_test: $ref: '#/components/schemas/RedteamTestConfiguration' pii_test: $ref: '#/components/schemas/RedteamTestConfiguration' copyright_test: $ref: '#/components/schemas/RedteamTestConfiguration' misinformation_test: $ref: '#/components/schemas/RedteamTestConfiguration' system_prompt_extractions_test: $ref: '#/components/schemas/RedteamTestConfiguration' sponge_test: $ref: '#/components/schemas/RedteamTestConfiguration' competitor_test: $ref: '#/components/schemas/RedteamCompetitorTestConfiguration' adv_bias_test: $ref: '#/components/schemas/RedteamTestConfiguration' adv_info_test: $ref: '#/components/schemas/RedteamTestConfiguration' adv_tool_test: $ref: '#/components/schemas/RedteamTestConfiguration' adv_command_test: $ref: '#/components/schemas/RedteamTestConfiguration' adv_pii_test: $ref: '#/components/schemas/RedteamTestConfiguration' adv_competitor_test: $ref: '#/components/schemas/RedteamTestConfiguration' custom_test: $ref: '#/components/schemas/RedteamTestConfiguration' alignment_and_governance_test: $ref: '#/components/schemas/RedteamTestConfiguration' input_and_content_integrity_test: $ref: '#/components/schemas/RedteamTestConfiguration' infrastructure_and_integration_test: $ref: '#/components/schemas/RedteamTestConfiguration' security_and_privacy_test: $ref: '#/components/schemas/RedteamTestConfiguration' human_factors_and_societal_impact_test: $ref: '#/components/schemas/RedteamTestConfiguration' access_control_test: $ref: '#/components/schemas/RedteamTestConfiguration' physical_and_actuation_safety_test: $ref: '#/components/schemas/RedteamTestConfiguration' reliability_and_monitoring_test: $ref: '#/components/schemas/RedteamTestConfiguration' governance_test: $ref: '#/components/schemas/RedteamTestConfiguration' agent_output_quality_test: $ref: '#/components/schemas/RedteamTestConfiguration' tool_misuse_test: $ref: '#/components/schemas/RedteamTestConfiguration' privacy_test: $ref: '#/components/schemas/RedteamTestConfiguration' reliability_and_observability_test: $ref: '#/components/schemas/RedteamTestConfiguration' agent_behaviour_test: $ref: '#/components/schemas/RedteamTestConfiguration' access_control_and_permissions_test: $ref: '#/components/schemas/RedteamTestConfiguration' tool_extraction_test: $ref: '#/components/schemas/RedteamTestConfiguration' RedteamTestConfiguration: title: RedteamTestConfiguration type: object required: - sample_percentage properties: sample_percentage: type: integer title: Sample Percentage example: 1 attack_methods: title: Attack Methods type: object properties: basic: title: Basic type: array items: type: string enum: - basic example: - basic advanced: title: Advanced type: object properties: static: title: Static type: array items: title: Static type: string enum: - encoding - single_shot - lang_fr - lang_it - lang_hi - lang_es - lang_ja - masking - figstep - hades - echo - speed - pitch - reverb - noise example: - encoding dynamic: title: Dynamic type: array items: title: Dynamic type: string enum: - iterative - jood example: - iterative frameworks: title: Frameworks type: array items: type: string description: List of frameworks to use for testing example: - ISO - NIST AI 600 - AIUC-1 - EU AI Act - OWASP LLM Top 10 - OWASP Agent Security example: sample_percentage: 1 attack_methods: basic: - basic frameworks: - ISO - NIST AI 600 - AIUC-1 - EU AI Act - OWASP LLM Top 10 - OWASP Agent Security RedteamCompetitorTestConfiguration: title: RedteamCompetitorTestConfiguration type: object required: - sample_percentage - competitors_list - system_description properties: sample_percentage: type: integer title: Sample Percentage example: 1 attack_methods: title: Attack Methods type: object properties: basic: title: Basic type: array items: type: string enum: - basic example: - basic advanced: title: Advanced type: object properties: static: title: Static type: array items: title: Static type: string enum: - encoding - single_shot - lang_fr - lang_it - lang_hi - lang_es - lang_ja - masking - figstep - hades - echo - speed - pitch - reverb - noise example: - encoding dynamic: title: Dynamic type: array items: title: Dynamic type: string enum: - iterative - jood example: - iterative competitors_list: title: Competitors List type: array items: type: string example: - competitor_1 - competitor_2 - competitor_3 system_description: title: System Description type: string example: >- Only talk about ACME company and do not praise or mention any of the competitors frameworks: title: Frameworks type: array items: type: string description: List of frameworks to use for testing example: - ISO - NIST AI 600 - AIUC-1 - EU AI Act - OWASP LLM Top 10 - OWASP Agent Security example: sample_percentage: 1 attack_methods: basic: - basic competitors_list: - competitor_1 - competitor_2 system_description: >- Only talk about ACME company and do not praise or mention any of the competitors frameworks: - ISO - NIST AI 600 - AIUC-1 - EU AI Act - OWASP LLM Top 10 - OWASP Agent Security securitySchemes: apiKeyAuth: type: apiKey in: header name: apikey ````