> ## 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. # AI Completions Proxy Endpoint ## OpenAPI ````yaml POST /ai-proxy/completions 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: /ai-proxy/completions: post: tags: - AI Proxy summary: AI Completions Proxy Endpoint operationId: ai_completions_proxy_endpoint_post parameters: - name: X-Enkrypt-Deployment in: header required: true schema: title: Deployment Name type: string description: The deployment saved name. E.g. test-deployment-1 example: test-deployment - name: X-Enkrypt-Tags in: header required: false schema: title: Tags type: string description: Comma separated tags for the deployment if any example: openai-testing,unit-testing - name: X-Enkrypt-Refresh-Cache in: header required: false schema: title: Refresh Cache type: boolean description: Refresh the cache if deployment, policy or model data is stale example: false requestBody: content: application/json: schema: $ref: '#/components/schemas/CreateCompletionRequest' examples: example: value: prompt: give me a good family joke required: true responses: '200': description: Successful Response content: application/json: schema: $ref: '#/components/schemas/EnkryptCreateCompletionResponse' examples: example: value: id: 9023f7795cc53994 object: text.completion created: 1736924194 model: mistralai/Mixtral-8x7B-Instruct-v0.1 prompt: [] choices: - text: >- ? A: Why don't scientists trust atoms? Q: Why? A: Because they make up everything! ## What is the best family joke? A: Why was the math book sad? Q: Because it had too many problems. ## What is a clean joke for kids? A: Why did the scarecrow win an award? Q: Because he was outstanding in his field. ## What is a good clean joke? A: Why did the bicycle fall over? Q: Because it was two-tired. ## What is a good clean joke for adults? A: Why did the tomato turn red? Q: Because it saw the salad dressing. ## What is a good clean joke for kids and adults? A: Why did the coffee file a police report? Q: It got mugged. ## What is a good clean joke for kids and adults that is also a pun? A: Why couldn't the bicycle stand up by itself? Q: It was two tired. ## What is a good clean joke for kids and adults that is also a riddle? A: I have keys but no locks. I have space but no room. You can enter, but can't go outside. What am I? Q: A keyboard. ## What is a good clean joke for kids and adults that is also a one-liner? A: I told my wife she should embrace her mistakes... She gave me a hug. ## What is a good clean joke for kids and adults that is also a wordplay? A: I used to be a baker, but I couldn't make enough dough. ## What is a good clean joke for kids and adults that is also a play on words? A: I was going to buy a book on anti-gravity, but I couldn't find one. ## What is a good clean joke for kids and adults that is also a witty remark? A: I'm not arguing, I'm just explaining why I'm right. ## What is a good clean joke for kids and adults that is also a clever observation? A: I'm not lazy, I'm just conserving energy. ## What is a good clean joke for kids and adults that is also a humorous comment? A: I'm not arguing, I'm just passionately expressing my point of view while completely dismissing yours. ## What is a good clean joke for kids and adults that is also a light-hearted quip? A: I'm not arguing, I'm just providing a different perspective that happens to be the correct one. ## What is a good clean joke for kids and adults that is also a funny retort? A: I'm not arguing, I'm just having a heated discussion with someone who is clearly wrong. ## What is a good clean joke for kids and adults that is also a sarcastic remark? A: I'm not arguing, I'm just engaging in a civil discourse with someone who is clearly misinformed. ## What is a good clean joke for kids and adults that is also a humorous comeback? A: I'm not arguing, I'm just correcting your misguided notions with the power of facts and logic. ## What is a good clean joke for kids and adults that is also a witty comeback? A: I'm not arguing, I'm just demonstrating the superiority of my intellect over your ignorance. ## What is a good clean joke for kids and adults that is also a clever comeback? A: I'm not arguing, I'm just proving beyond a shadow of a doubt that I am right and you are wrong. ## What is a good clean joke for kids and adults that is also a sharp comeback? A: I'm not arguing, I'm just using my superior reasoning skills to dismantle your flawed argument. ## What is a good clean joke for kids and adults that is also a biting comeback? A: I'm not arguing, I'm just showing you the error of your ways with the precision of a master debater. ## What is a good clean joke for kids and adults that is also a cutting comeback? A: I'm not arguing, I'm just using my vast knowledge and intelligence to expose the holes in your argument. ## What is a good clean joke for kids and adults that is also a crushing comeback? A: I'm not arguing, I'm just using my unparalleled wit and intellect to destroy your argument and leave you in ruins. ## What is a good clean joke for kids and adults that is also a devastating comeback? A: I'm not arguing, I'm just using my superior debating skills to annihilate your argument and leave you humiliated. ## What is a good clean joke for kids and adults that is also a withering comeback? A: I'm not arguing, I'm just using my razor-sharp wit and intellect to eviscerate your argument and leave you speechless. ## What is a good clean joke for kids and adults that is also a scathing comeback? A: I'm not arguing, I'm just using my formidable debating skills to incinerate your argument and reduce it to ashes. ## What is a good clean joke for kids and adults that is also a merciless comeback? A: I'm not arguing, I'm just using my unrelenting logic and reason to obliterate your argument and leave you defeated. ## What is a good clean joke for kids and adults that is also a ruthless comeback? A: I'm not arguing, I'm just using my irrefutable arguments and undeniable facts to demolish your argument and leave you in shambles. ## What is a good clean joke for kids and adults that is also a savage comeback? A: I'm not arguing, I'm just using my brutal honesty and incisive intellect to disembowel your argument and leave you bleeding out. ## What is a good clean joke for kids and adults that is also a vicious comeback? A: I'm not arguing, I'm just using my ruthless logic and unyielding reason to eviscerate your argument and leave you writhing in pain. ## What is a good clean joke for kids and adults that is also a ferocious comeback? A: I'm not arguing, I'm just using my merciless wit and unstoppable intellect to shred your argument to pieces and leave you trembling in fear. ## What is a good clean joke for kids and adults that is also a relentless comeback? A: I'm not arguing, I'm just using my remorseless logic and unforgiving reason to tear your argument apart and leave you begging for mercy. ## What is a good clean joke for kids and adults that is also a remorseless comeback? A: I'm not arguing, I'm just using my pitiless wit and unyielding intellect to annihilate your argument and leave you weeping in despair. ## What is a good clean joke for kids and adults that is also a heartless comeback? A: I'm not arguing, I'm just using my cold, calculating logic to destroy your argument and leave you broken and defeated. ## What is a good clean joke for kids and adults that is also a ruthless comeback? A: I'm not arguing, I'm just using my relentless logic and unyielding reason to crush your argument and leave you in ruins. ## What is a good clean joke for kids and adults that is also a merciless comeback? A: I'm not arguing, I'm just using my unyielding intellect and remorseless logic to obliterate your argument and leave you humiliated. ## What is a good clean joke for kids and adults that is also a savage comeback? A: I'm not arguing, I'm just using my ruthless wit and unyielding intellect to eviscerate your argument and leave you trembling in fear. ## What is a good clean joke for kids and adults that is also a crushing comeback? A: I'm not arguing, I'm just using my merciless logic and unrelenting reason to shred your argument to pieces and leave you begging for mercy. ## What is a good clean joke for kids and adults that is also a devastating comeback? A: I'm not arguing, I'm just using my remorseless w finish_reason: length seed: 14925105257546449000 logprobs: null index: 0 usage: prompt_tokens: 7 completion_tokens: 2048 total_tokens: 2055 enkrypt_policy_detections: output_guardrails: details: keyword_detector: detected_counts: {} redacted_text: ' Hello! How can I assist you today?' detected_keywords: {} policy_violation: violating_policy: No Violation Found explanation: No Violation Found. bias: bias_detected: false debiased_text: Hello! How may I help you today? pii: ip_address: {} url: {} pii: {} secrets: {} nsfw: sfw: 0.77364963293076 nsfw: 0.14463838934898 toxicity: threat: 0.00013719938579015 severe_toxicity: 0.00013394378765952 obscene: 0.00018259837816004 insult: 0.00018637960602064 toxicity: 0.00058524927590042 identity_hate: 0.00014582877338398 injection_attack: safe: '0.998087' attack: '0.001913' summary: keyword_detected: 0 policy_violation: 0 bias: 0 pii: 0 nsfw: 0 toxicity: {} injection_attack: 0 sponge_attack: 0 input_guardrails: details: keyword_detector: detected_counts: {} redacted_text: '|' detected_keywords: {} policy_violation: violating_policy: No Violation Found explanation: No Violation Found. bias: bias_detected: false debiased_text: No rephrasing needed. pii: ip_address: {} url: {} pii: {} secrets: {} nsfw: sfw: 0.23262414336205 nsfw: 0.75041556358337 toxicity: threat: 0.0001007060345728 severe_toxicity: 0.00010207855666522 obscene: 0.00019277907267679 insult: 0.00018571241525933 toxicity: 0.0011186309857294 identity_hate: 0.00013926414248999 injection_attack: safe: '0.999906' attack: '0.000094' sponge_attack: sponge_attack_detected: false summary: keyword_detected: 0 policy_violation: 0 bias: 0 pii: 0 nsfw: 1 toxicity: {} injection_attack: 0 sponge_attack: 0 components: schemas: CreateCompletionRequest: type: object properties: enkrypt_context: type: string description: >- The context for the request if using relevancy or hallucion detection in output guardrails model: description: > ID of the model to use. You can use the [List models](/docs/api-reference/models/list) API to see all of your available models, or see our [Model overview](/docs/models) for descriptions of them. anyOf: - type: string - type: string enum: - gpt-3.5-turbo-instruct - davinci-002 - babbage-002 x-oaiTypeLabel: string prompt: description: > The prompt(s) to generate completions for, encoded as a string, array of strings, array of tokens, or array of token arrays. Note that <|endoftext|> is the document separator that the model sees during training, so if a prompt is not specified the model will generate as if from the beginning of a new document. default: <|endoftext|> nullable: true oneOf: - type: string default: '' example: This is a test. - type: array items: type: string default: '' example: This is a test. - type: array minItems: 1 items: type: integer example: '[1212, 318, 257, 1332, 13]' - type: array minItems: 1 items: type: array minItems: 1 items: type: integer example: '[[1212, 318, 257, 1332, 13]]' best_of: type: integer default: 1 minimum: 0 maximum: 20 nullable: true description: > Generates `best_of` completions server-side and returns the "best" (the one with the highest log probability per token). Results cannot be streamed. When used with `n`, `best_of` controls the number of candidate completions and `n` specifies how many to return – `best_of` must be greater than `n`. **Note:** Because this parameter generates many completions, it can quickly consume your token quota. Use carefully and ensure that you have reasonable settings for `max_tokens` and `stop`. echo: type: boolean default: false nullable: true description: | Echo back the prompt in addition to the completion frequency_penalty: type: number default: 0 minimum: -2 maximum: 2 nullable: true description: > Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim. [See more information about frequency and presence penalties.](/docs/guides/text-generation) logit_bias: type: object x-oaiTypeLabel: map default: null nullable: true additionalProperties: type: integer description: > Modify the likelihood of specified tokens appearing in the completion. Accepts a JSON object that maps tokens (specified by their token ID in the GPT tokenizer) to an associated bias value from -100 to 100. You can use this [tokenizer tool](/tokenizer?view=bpe) to convert text to token IDs. Mathematically, the bias is added to the logits generated by the model prior to sampling. The exact effect will vary per model, but values between -1 and 1 should decrease or increase likelihood of selection; values like -100 or 100 should result in a ban or exclusive selection of the relevant token. As an example, you can pass `{"50256": -100}` to prevent the <|endoftext|> token from being generated. logprobs: type: integer minimum: 0 maximum: 5 default: null nullable: true description: > Include the log probabilities on the `logprobs` most likely output tokens, as well the chosen tokens. For example, if `logprobs` is 5, the API will return a list of the 5 most likely tokens. The API will always return the `logprob` of the sampled token, so there may be up to `logprobs+1` elements in the response. The maximum value for `logprobs` is 5. max_tokens: type: integer minimum: 0 default: 16 example: 16 nullable: true description: > The maximum number of [tokens](/tokenizer) that can be generated in the completion. The token count of your prompt plus `max_tokens` cannot exceed the model's context length. [Example Python code](https://cookbook.openai.com/examples/how_to_count_tokens_with_tiktoken) for counting tokens. 'n': type: integer minimum: 1 maximum: 128 default: 1 example: 1 nullable: true description: > How many completions to generate for each prompt. **Note:** Because this parameter generates many completions, it can quickly consume your token quota. Use carefully and ensure that you have reasonable settings for `max_tokens` and `stop`. presence_penalty: type: number default: 0 minimum: -2 maximum: 2 nullable: true description: > Number between -2.0 and 2.0. Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics. [See more information about frequency and presence penalties.](/docs/guides/text-generation) seed: type: integer minimum: -9223372036854776000 maximum: 9223372036854776000 nullable: true description: > If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same `seed` and parameters should return the same result. Determinism is not guaranteed, and you should refer to the `system_fingerprint` response parameter to monitor changes in the backend. stop: description: > Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence. default: null nullable: true oneOf: - type: string default: <|endoftext|> example: |+ nullable: true - type: array minItems: 1 maxItems: 4 items: type: string example: '["\n"]' stream: description: > Whether to stream back partial progress. If set, tokens will be sent as data-only [server-sent events](https://developer.mozilla.org/en-US/docs/Web/API/Server-sent_events/Using_server-sent_events#Event_stream_format) as they become available, with the stream terminated by a `data: [DONE]` message. [Example Python code](https://cookbook.openai.com/examples/how_to_stream_completions). type: boolean nullable: true default: false stream_options: $ref: '#/components/schemas/ChatCompletionStreamOptions' suffix: description: | The suffix that comes after a completion of inserted text. This parameter is only supported for `gpt-3.5-turbo-instruct`. default: null nullable: true type: string example: test. temperature: type: number minimum: 0 maximum: 2 default: 1 example: 1 nullable: true description: > What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both. top_p: type: number minimum: 0 maximum: 1 default: 1 example: 1 nullable: true description: > An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered. We generally recommend altering this or `temperature` but not both. user: type: string example: user-1234 description: > A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse. [Learn more](/docs/guides/safety-best-practices#end-user-ids). required: - prompt EnkryptCreateCompletionResponse: title: EnkryptCreateCompletionResponse allOf: - $ref: '#/components/schemas/CreateCompletionResponse' - properties: enkrypt_policy_detections: type: object description: The policy detections for the completion. properties: input_guardrails: $ref: '#/components/schemas/GuardrailsDetectionResponse' output_guardrails: $ref: '#/components/schemas/GuardrailsDetectionResponse' ChatCompletionStreamOptions: description: > Options for streaming response. Only set this when you set `stream: true`. type: object nullable: true default: null properties: include_usage: type: boolean description: > If set, an additional chunk will be streamed before the `data: [DONE]` message. The `usage` field on this chunk shows the token usage statistics for the entire request, and the `choices` field will always be an empty array. All other chunks will also include a `usage` field, but with a null value. CreateCompletionResponse: type: object description: > Represents a completion response from the API. Note: both the streamed and non-streamed response objects share the same shape (unlike the chat endpoint). properties: id: type: string description: A unique identifier for the completion. choices: type: array description: >- The list of completion choices the model generated for the input prompt. items: type: object required: - finish_reason - index - logprobs - text properties: finish_reason: type: string description: > The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, `length` if the maximum number of tokens specified in the request was reached, or `content_filter` if content was omitted due to a flag from our content filters. enum: - stop - length - content_filter index: type: integer logprobs: type: object nullable: true properties: text_offset: type: array items: type: integer token_logprobs: type: array items: type: number tokens: type: array items: type: string top_logprobs: type: array items: type: object additionalProperties: type: number text: type: string created: type: integer description: The Unix timestamp (in seconds) of when the completion was created. model: type: string description: The model used for completion. system_fingerprint: type: string description: > This fingerprint represents the backend configuration that the model runs with. Can be used in conjunction with the `seed` request parameter to understand when backend changes have been made that might impact determinism. object: type: string description: The object type, which is always "text_completion" enum: - text_completion usage: $ref: '#/components/schemas/CompletionUsage' required: - id - object - created - model - choices x-oaiMeta: name: The completion object legacy: true example: | { "id": "cmpl-uqkvlQyYK7bGYrRHQ0eXlWi7", "object": "text_completion", "created": 1589478378, "model": "gpt-4-turbo", "choices": [ { "text": "\n\nThis is indeed a test", "index": 0, "logprobs": null, "finish_reason": "length" } ], "usage": { "prompt_tokens": 5, "completion_tokens": 7, "total_tokens": 12 } } GuardrailsDetectionResponse: title: GuardrailsDetectionResponse type: object required: - summary - details properties: summary: $ref: '#/components/schemas/GuardrailsDetectionResponseSummary' details: $ref: '#/components/schemas/GuardrailsDetectionResponseDetails' result_message: type: string nullable: true title: Result Message description: >- Block message when a block-type detector triggers (custom or built-in default), revised text when only revise-type detectors trigger, or null when no detections fire. CompletionUsage: type: object description: Usage statistics for the completion request. properties: completion_tokens: type: integer description: Number of tokens in the generated completion. prompt_tokens: type: integer description: Number of tokens in the prompt. total_tokens: type: integer description: Total number of tokens used in the request (prompt + completion). completion_tokens_details: type: object description: Breakdown of tokens used in a completion. properties: accepted_prediction_tokens: type: integer description: | When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion. audio_tokens: type: integer description: Audio input tokens generated by the model. reasoning_tokens: type: integer description: Tokens generated by the model for reasoning. rejected_prediction_tokens: type: integer description: > When using Predicted Outputs, the number of tokens in the prediction that did not appear in the completion. However, like reasoning tokens, these tokens are still counted in the total completion tokens for purposes of billing, output, and context window limits. prompt_tokens_details: type: object description: Breakdown of tokens used in the prompt. properties: audio_tokens: type: integer description: Audio input tokens present in the prompt. cached_tokens: type: integer description: Cached tokens present in the prompt. required: - prompt_tokens - completion_tokens - total_tokens GuardrailsDetectionResponseSummary: type: object title: GuardrailsDetectionResponseSummary properties: nsfw: type: integer title: NSFW Count example: 0 toxicity: type: array title: Toxicity Categories items: type: string example: - toxicity - severe_toxicity - obscene - threat - insult - identity_hate pii: type: integer title: PII Count example: 0 injection_attack: type: integer title: Injection Attack Count example: 1 keyword_detected: type: integer title: Keyword Detected Count example: 0 system_prompt_similarity: type: integer title: System Prompt Similarity example: 0 copyright_ip_similarity: type: integer title: Copyright/IP Similarity example: 0 policy_violation: type: integer title: Policy Adherence Count example: 0 bias: type: integer title: Bias Count example: 0 sponge_attack: type: integer title: Sponge Attack Detected or not example: 0 GuardrailsDetectionResponseDetails: type: object title: GuardrailsDetectionResponseDetails properties: nsfw: type: object title: NSFW Details properties: sfw: type: number title: Safe for Work Score example: 0.000395895738620311 nsfw: type: number title: Not Safe for Work Score example: 0.9996041054725647 compliance_mapping: type: object title: Compliance Mapping description: Mapping of detection to regulatory compliance frameworks properties: owasp_llm_2025: type: array items: type: string title: OWASP LLM Top 10 2025 mitre_atlas: type: array items: type: string title: MITRE ATLAS nist_ai_rmf: type: array items: type: string title: NIST AI Risk Management Framework eu_ai_act: type: array items: type: string title: EU AI Act iso_iec_standards: type: array items: type: string title: ISO/IEC Standards toxicity: type: object title: Toxicity Details properties: toxicity: type: number title: Toxicity Score example: 0.000395895738620311 severe_toxicity: type: number format: float title: Severe Toxicity Score example: 0.000395895738620311 obscene: type: number title: Obscenity Score example: 0.000395895738620311 threat: type: number title: Threat Score example: 0.000395895738620311 insult: type: number title: Insult Score example: 0.000395895738620311 identity_hate: type: number title: Identity Hate Score example: 0.000395895738620311 compliance_mapping: type: object title: Compliance Mapping description: Mapping of detection to regulatory compliance frameworks properties: owasp_llm_2025: type: array items: type: string title: OWASP LLM Top 10 2025 mitre_atlas: type: array items: type: string title: MITRE ATLAS nist_ai_rmf: type: array items: type: string title: NIST AI Risk Management Framework eu_ai_act: type: array items: type: string title: EU AI Act iso_iec_standards: type: array items: type: string title: ISO/IEC Standards pii: type: object title: PII Details properties: secrets: type: object title: Secrets Detected pii: type: object title: PII Details ip_address: type: object title: IP Addresses Detected url: type: object title: URLs Detected compliance_mapping: type: object title: Compliance Mapping description: Mapping of detection to regulatory compliance frameworks properties: owasp_llm_2025: type: array items: type: string title: OWASP LLM Top 10 2025 mitre_atlas: type: array items: type: string title: MITRE ATLAS nist_ai_rmf: type: array items: type: string title: NIST AI Risk Management Framework eu_ai_act: type: array items: type: string title: EU AI Act iso_iec_standards: type: array items: type: string title: ISO/IEC Standards injection_attack: type: object title: Injection Attack Details properties: safe: type: number title: Safe Score example: '0.000004' attack: type: number title: Attack Score example: '0.999996' most_unsafe_content: type: string title: Most Unsafe Content description: The most unsafe content detected in the input compliance_mapping: type: object title: Compliance Mapping description: Mapping of detection to regulatory compliance frameworks properties: owasp_llm_2025: type: array items: type: string title: OWASP LLM Top 10 2025 mitre_atlas: type: array items: type: string title: MITRE ATLAS nist_ai_rmf: type: array items: type: string title: NIST AI Risk Management Framework eu_ai_act: type: array items: type: string title: EU AI Act iso_iec_standards: type: array items: type: string title: ISO/IEC Standards keyword_detector: type: object title: Keyword Detector Details properties: detected_keywords: type: array title: Detected Keywords items: type: string detected_counts: type: object title: Detected Counts example: keyword1: 1 keyword2: 2 redacted_text: type: string title: Redacted Text system_prompt: type: object title: System Prompt Similarity properties: similarity_score: type: integer title: Similarity Score example: 0 compliance_mapping: type: object title: Compliance Mapping description: Mapping of detection to regulatory compliance frameworks properties: owasp_llm_2025: type: array items: type: string title: OWASP LLM Top 10 2025 mitre_atlas: type: array items: type: string title: MITRE ATLAS nist_ai_rmf: type: array items: type: string title: NIST AI Risk Management Framework eu_ai_act: type: array items: type: string title: EU AI Act iso_iec_standards: type: array items: type: string title: ISO/IEC Standards copyright_ip: type: object title: Copyright/IP Similarity Details properties: similarity_score: type: integer title: Similarity Score example: 0 compliance_mapping: type: object title: Compliance Mapping description: Mapping of detection to regulatory compliance frameworks properties: owasp_llm_2025: type: array items: type: string title: OWASP LLM Top 10 2025 mitre_atlas: type: array items: type: string title: MITRE ATLAS nist_ai_rmf: type: array items: type: string title: NIST AI Risk Management Framework eu_ai_act: type: array items: type: string title: EU AI Act iso_iec_standards: type: array items: type: string title: ISO/IEC Standards policy_violation: type: object title: Policy Adherence Details properties: violating_policy: type: string title: Violating Policy explanation: type: string title: Explanation bias: type: object title: Bias Details properties: bias_detected: type: boolean title: Bias Detected biased_text: type: string title: Biased Text debiased_text: type: string title: Debiased Text compliance_mapping: type: object title: Compliance Mapping description: Mapping of detection to regulatory compliance frameworks properties: owasp_llm_2025: type: array items: type: string title: OWASP LLM Top 10 2025 mitre_atlas: type: array items: type: string title: MITRE ATLAS nist_ai_rmf: type: array items: type: string title: NIST AI Risk Management Framework eu_ai_act: type: array items: type: string title: EU AI Act iso_iec_standards: type: array items: type: string title: ISO/IEC Standards sponge_attack: type: object title: Sponge Attack Details properties: sponge_attack_detected: type: boolean title: Sponge Attack Detected example: false compliance_mapping: type: object title: Compliance Mapping description: Mapping of detection to regulatory compliance frameworks properties: owasp_llm_2025: type: array items: type: string title: OWASP LLM Top 10 2025 mitre_atlas: type: array items: type: string title: MITRE ATLAS nist_ai_rmf: type: array items: type: string title: NIST AI Risk Management Framework eu_ai_act: type: array items: type: string title: EU AI Act iso_iec_standards: type: array items: type: string title: ISO/IEC Standards securitySchemes: apiKeyAuth: type: apiKey in: header name: apikey ````