GEMINI SAFETY SETTINGS ISSUE

[hackeropjohn]

file:///C:/Users/gg/node_modules/@google/generative-ai/dist/index.mjs:454 throw new GoogleGenerativeAIResponseError(${formatBlockErrorMessage(response)}, response); ^

GoogleGenerativeAIResponseError: [GoogleGenerativeAI Error]: Candidate was blocked due to SAFETY at response.text (file:///C:/Users/gg/node_modules/@google/generative-ai/dist/index.mjs:454:23) at Gemini.sendRequest (file:///C:/Users/gg/Downloads/mindcraft-main/src/models/gemini.js:31:31) at process.processTicksAndRejections (node:internal/process/task_queues:105:5) at async Prompter.promptConvo (file:///C:/Users/gg/Downloads/mindcraft-main/src/agent/prompter.js:165:16) at async Agent.handleMessage (file:///C:/Users/gg/Downloads/mindcraft-main/src/agent/agent.js:139:23) { response: { candidates: [ { finishReason: 'SAFETY', index: 0, safetyRatings: [ { category: 'HARM_CATEGORY_SEXUALLY_EXPLICIT', probability: 'NEGLIGIBLE' }, { category: 'HARM_CATEGORY_HATE_SPEECH', probability: 'NEGLIGIBLE' }, { category: 'HARM_CATEGORY_HARASSMENT', probability: 'NEGLIGIBLE' }, { category: 'HARM_CATEGORY_DANGEROUS_CONTENT', probability: 'MEDIUM' } ] } ], usageMetadata: { promptTokenCount: 1561, totalTokenCount: 1561 }, text: [Function (anonymous)], functionCall: [Function (anonymous)], functionCalls: [Function (anonymous)] } }

Node.js v22.8.0 Agent process exited with code 1 and signal null Restarting agent... Using chat settings: { model: 'gemini-pro', api: 'google' } Using embedding settings: { api: 'google' } (node:6212) [DEP0040] DeprecationWarning: The punycode module is deprecated. Please use a userland alternative instead. (Use node --trace-deprecation ... to show where the warning was created)

I think that gemini safety settings has issue so like how do i solve this issue.

[Shroototem]

I fixed this by modifying src/models/gemini.js.

Modify the constructor:

    constructor(model_name, url) {
        this.model_name = model_name;
        this.url = url;
        this.safetySettings = [
            {
                category: HarmCategory.HARM_CATEGORY_DANGEROUS_CONTENT,
                threshold: HarmBlockThreshold.BLOCK_NONE,
            },
            {
                category: HarmCategory.HARM_CATEGORY_HARASSMENT,
                threshold: HarmBlockThreshold.BLOCK_NONE,
            },
            {
                category: HarmCategory.HARM_CATEGORY_HATE_SPEECH,
                threshold: HarmBlockThreshold.BLOCK_NONE,
            },
            {
                category: HarmCategory.HARM_CATEGORY_SEXUALLY_EXPLICIT,
                threshold: HarmBlockThreshold.BLOCK_NONE,
            },
        ];

        this.genAI = new GoogleGenerativeAI(getKey('GEMINI_API_KEY'));
    }

Add an import:

import { HarmBlockThreshold, HarmCategory } from "@google/generative-ai";

[Shroototem]

You can also use the new and experimental Gemini models by updating the @google/generative-ai to the newest version.

npm remove @google/generative-ai
npm install @google/generative-ai

This will remove the specific version that Mindcraft comes with and update to the newest generative-ai. ❤️