Is CAI Tools safe?

// Get prompt, import char settings, import persona, import special instructionslet prompt = defaultGreetingGenPrompt_GPT;let rawPersonaText = persona ? '\n\n' + Object.keys(persona)    .filter((key) => persona[key]) // Leave out the empty and undefined    .map((key) => `${key}: ${persona[key]}`).join('\n') : '';prompt = prompt.replace(/{{dynamic-char-settings/g, rawCharPersonalityText)    .replace(/{{user-persona}}/g, rawPersonaText)    .replace(/{{special-instructions}}/g, special ? `\n\n${special}` : '');this.ReportProgress("Loading... (Asking ChatGPT)");generateGreetingWithChatGPT({ prompt, model: this.model_select.value })

export async function generateGreetingWithChatGPT(args) {    try {        const gptLoginData = await getChatGPTAccessToken();        if (!gptLoginData.token)            return { error: "not_logged_in", message: `You need to login to your ChatGPT account to be able to use this feature.<br/><a href='https://chatgpt.com/' target='_blank'>https://chatgpt.com/</a><br/>If you are already logged in, try logging out & in on ChatGPT.${gptLoginData.error ? `<br/>${gptLoginData.error}` : ''}` };        // Ask gpt        const res = await askChatGPT(args.prompt, gptLoginData.token, args.model);        return res;    } catch (error) {        return { error: "unexpected_error", message: error.message || error };    }}// Get session for access tokenasync function getChatGPTAccessToken() {    try {        const res_session = await fetch("https://chatgpt.com/api/auth/session");        if (!res_session.ok)            throw new Error("ChatGPT login error.", res_session.status, res_session.statusText)        const data_session = await res_session.json();        return { token: data_session.accessToken };    } catch (error) {        return { error };    }}

async function askChatGPT(prompt, chatgptAccessToken, model, maxTries = 3, modelSelection = 1) {    const randomDeviceId = crypto.randomUUID();    try {        // For chat requirements        const chatReqToken = await solveChallenge("chat_req_token", { chatReqSeed: "" + Math.random(), chatReqDifficulty: "0" });        if (!chatReqToken) {            return { error: "requirements_failed", message: "Couldn't generate the first gpt chat requirement token." };        }        const res_req = await fetch("https://chatgpt.com/backend-api/sentinel/chat-requirements", {            "headers": {                "accept": "*/*",                "authorization": `Bearer ${chatgptAccessToken}`,                "content-type": "application/json"            },            "body": JSON.stringify({ p: chatReqToken }),            "method": "POST"        });        if (!res_req.ok) {            try {                const errorData = await res_req.json();                if (errorData.detail?.code === "token_expired") {                    return { error: "token_expired", message: "ChatGPT session has expired. Please login again.<br/><a href='https://chatgpt.com/' target='_blank'>https://chatgpt.com/</a>" }                }                throw "";            } catch (err) {                throw new Error("Error while trying to get ChatGPT chat requirements.", res_req.status, res_req.statusText);            }        }        const data_req = await res_req.json();        const chatReqSeed = data_req.proofofwork.seed;        const chatReqDifficulty = data_req.proofofwork.difficulty;        const chatReqTokenFinal = data_req.token;        if (!chatReqTokenFinal) {            return { error: "requirements_failed", message: "Couldn't generate the final gpt chat requirement token." };        }        let chatReqArkoseToken; // required for gpt-4 but I am not using gpt-4 even when available        // Generate proof token        const proofToken = await solveChallenge("proof_token", { chatReqSeed, chatReqDifficulty });        if (!chatReqTokenFinal) {            return { error: "requirements_failed", message: "Couldn't generate the proof token." };        }        // const model = modelSelection === 1 ? await getChatGPTModel(chatgptAccessToken) : "text-davinci-002-render-sha";        const payload = {            "action": "next",            "arkose_token": chatReqArkoseToken,            "messages": [                {                    "id": crypto.randomUUID(),                    "author": {                        "role": "user"                    },                    "content": {                        "content_type": "text",                        "parts": [                            prompt                        ]                    }                }            ],            "model": model || "text-davinci-002-render-sha",            "parent_message_id": crypto.randomUUID(),            "is_visible": false, // false to skip extra request to make it invisible            "force_nulligen": false,            "force_paragen": false,            "force_paragen_model_slug": "",            "force_rate_limit": false,            "suggestions": [],            "conversation_mode": {                "kind": "primary_assistant"            }        }        const headers = {            "accept": "text/event-stream",            "content-type": "application/json",            "oai-device-id": randomDeviceId,            "oai-language": "en-us",            "authorization": `Bearer ${chatgptAccessToken}`,            "openai-sentinel-chat-requirements-token": chatReqTokenFinal,            ...(proofToken ? { "openai-sentinel-proof-token": proofToken } : {}),            ...(chatReqArkoseToken ? { "openai-sentinel-arkose-token": chatReqArkoseToken } : {})        }        // Start conversation        const res_convo = await fetch("https://chatgpt.com/backend-api/conversation", {            "headers": headers,            "body": JSON.stringify(payload),            "method": "POST"        });        if (!res_convo.ok) {            if (modelSelection !== 3) {                return await askChatGPT(prompt, chatgptAccessToken, model, maxTries, modelSelection + 1);            }            if (res_convo.status === 429) {                // "You've reached our limit of messages per hour. Please try again later."                return { error: "chatgpt_limit", message: "ChatGPT message limit for this model has been reached." };            }            if (maxTries > 0) {                maxTries--;                return await askChatGPT(prompt, chatgptAccessToken, model, maxTries, modelSelection);            } else {                throw new Error("Error while asking ChatGPT. Try logging out and in on chatgpt. " + res_convo.status + ", " + res_convo.statusText);            }        }        // Read the event stream        let finalItem = await readAllChunks(res_convo);        if ((!finalItem || !finalItem.gptResponse) && maxTries < 3) {            maxTries--;            return await askChatGPT(prompt, chatgptAccessToken, model, maxTries, modelSelection);        }        if (finalItem.error) {            return { error: "asking_failed", message: finalItem.error };        }        else if (!finalItem || !finalItem.gptResponse) {            return { error: "asking_failed", message: "Unknown error. No response returned." };        }        // Return the final result        return finalItem;    } catch (error) {        return { error: "unexpected_error", message: error.message || error };    }}