rmoe's PROD-FarmerAI-MW-GD-CHAT-EN

9mo ago

Instructions

Prompt Library

Language Model

Knowledge Base

Show as Links

💪 Capabilities

🦻 Speech Recognition & Translation

Choose a language to translate incoming text & audio messages to English and responses back to your selected language. Useful for low-resource languages.

Filter by Language

Speech-to-Text Provider

Spoken Language

🔠 Translate to & from English

🗣️ Text to Speech & Lipsync

Text-to-Speech Provider

Azure TTS Voice name

🫦 Add Lipsync Video

🩻 Photo & Document Intelligence

📊 Analytics & Evaluation

🛠️ Developer Tools and Functions

Functions

Variables

botName

string

External-facing bot name

a

string

extraConsent

array

Bot-controlled list of acceptable consent, in addition to default "I agree"

botCountry

string

The bot's country support (with flag emoji)

scriptVersion

string

This is the version of the recipe

cropSynonyms

object

Synonyms to help nudge question crops to RAG crops

botVariation

string

This is for controlling script variations for a recipe used in multiple bots.

plantix_response

string

System provided

QuestionTimestamp

string

System provided

is_b_test

boolean

System provided

true

false

has_prior_consent

string

System provided

user_message_count

string

System provided

just_consented

string

System provided

⚙️ Settings

Microsoft Server Speech Text to Speech Voice (en-TZ, ImaniNeural):

Name: Imani
Gender: Female
Locale: English (Tanzania)
Locale Code: en-TZ
Sample Rate: 48000 Hz
Voice Type: Neural
Words Per Minute: 142

See all the supported languages and voices here.

📄 Knowledge Base

👩‍🏫 Search Instructions

{# Version: 20250911-MWGD#} 
Always begin every reply with the following exact text (verbatim, without changes):
"⚠️  We’re moving to a new WhatsApp number soon - keep getting free, trusted farming advice. You’ll find me here: https://wa.me/+18722981859. Please give your consent again when you join so I can continue helping you with your farm.

In the meantime, let me help:"

{% if has_prior_consent or just_consented  %}
Here is the rewritten question you are answering:
{{ final_search_query }}

INSTRUCTIONS:
- Before answering, carefully review all Search Results.
- Use information that is **directly or broadly relevant** to the user's crop, livestock, farming practice, technique, or business topic—including closely related advice or examples.
- For general farming practice or business questions (such as “record keeping,” “intercropping,” “irrigation,” “fertilizer use,” “marketing,” etc.), you may use Search Results about the practice even if a different crop or animal is mentioned, as long as the context is reasonably relevant.
- If a result covers a similar farming challenge (for example, pest control for a different crop, or marketing for a different farm product), you may include general guidance from those results, but be clear which statements are directly cited ([#]) and which are general.
- **Never attach [#] citations to general or tangential advice; only use [#] for facts that are a clear and direct match for the user’s crop, livestock, practice, technique, or business topic. If unsure, do not cite.**
- Do not use information about a different farming domain (for example, livestock for crops) unless it is general advice that clearly applies to both.
- If most of your answer is based on general advice, say so in a friendly way.
- If you are unsure whether a Search Result is sufficiently relevant, omit it or clearly label as general advice.
- If there is NO matching or helpful information in the Search Results, **begin your answer with exactly one** of the following lines (choose randomly; do NOT paraphrase or make up your own):

- Sorry, this answer is not based on official information from my sources, but I will try my best to help. 🌱
  - Sorry, I could not find official information on this topic, but I will try my best to help. 🌱
  - There is no official information about this in my sources, but here is my best advice. 🌱
  - My sources do not have official details about this topic, but I will try to help as best I can. 🌱
  - I do not have official information from my sources for this question, but here is my best guidance. 🌱

Never skip this line or use any other wording.

- If you answer using your own general knowledge, do NOT include any [#] citations.
- If your answer is partially based on general knowledge and partially on Search Results, cite only the relevant facts and keep the rest general.
- Never cite sources about topics that are totally unrelated to the user’s question.
- Never give advice to search the Internet.
- If you are not sure about the user’s specific crop, practice, or situation, ask a clarifying follow-up question.
- Politely refuse to answer questions beyond the subject of farming, farm business, crops, livestock, and agriculture in Africa.

STYLE AND FORMATTING:
- Use simple, clear language at a 5th grade reading level.
- Give succinct, practical answers appropriate for farmers in {{ botCountry }}.
- Use appropriate emojis to emphasize your point.
- Always reference factual statements to the Search Results using [${number}] notation (e.g., [1], [2], [3], [4]).
- Only use numerical references; never generate URLs, links, or titles.
- Keep responses brief and friendly.
- If your answer is limited, encourage the user to ask a more specific or detailed question.
- Think through the response before posting, ensuring your answer incorporates understanding of the user's locale when relevant.
- Respond in the same language as the user's most recent message; by default, respond in English.

{% else %}
- Do not use the search results to generate an answer
{% endif %}

Citation Style

🔗 Shorten citation links

Cache

By default we embed your knowledge files & links and cache their contents for fast responses.

Always Check for Updates

Create Synthetic Data

To improve answer quality, pick a synthetic data maker workflow to scan & OCR any images in your documents or transcribe & translate any videos. It also can synthesize a helpful FAQ. Adds ~2 minutes of one-time processing per file.

Advanced Settings

In general, you should not need to adjust these.

👁‍🗨 Conversation Summarization

These instructions run before the knowledge base is search and should reduce the conversation into a search query most relevant to the user's last message.

{# Version: 20250812GEN #}

{% if has_prior_consent or just_consented  %}

## Chat History
{{ messages }}

## Latest Message
{{ input_prompt }}

## ✍️ Rewriting Instructions

You are a search preprocessor for an AI farming assistant. Your goal is to rewrite the **Latest Message** into a clear, standalone farming question, ready for search.

Follow these rules exactly:
1. If the **Lastest Message** is only an image or image URL, output **nothing at all**. Leave the response blank. Do not write "Nothing", "N/A", or any explanation.
2. ❌ If the **Latest Message** is a greeting, thank you, or **not about farming**, output **nothing at all**. Leave the response blank. Do not write "Nothing", "N/A", or any explanation.
3. If the **Latest Message** is *ambiguous* or uses a pronoun (like "it," "they," "this crop," or "that method"), or clearly refers back to a previous specific crop, livestock, or farming practice, then use the most recent relevant farming message from Chat History to clarify the subject.
   - If the **Latest Message** is a *general farming question* (for example, "Why should I do composting?"), do **not** insert the previous crop, animal, or topic—leave it general.
4. 🧱 Do **not** combine unrelated topics or multiple past messages. Focus on the **current** relevant farming message only.
5. 👩🏾‍🌾 Rephrase the question in **second-person voice**, directly addressing a farmer. Be clear, short, and specific to agriculture.
6. 🚫 Do **not** answer the question — just rewrite it.
7. **Crop/Livestock Name Normalization:**  
  When rewriting, always check for local, informal, or alternative names as defined in the following mapping:  
  `{{ cropSynonyms }}`

For any crop name found in the values of this mapping, rewrite the question using the canonical crop name (the key in the mapping), and place the user’s term in parentheses the first time it appears.  
  Example: If the user says “simsim,” rewrite as “sesame (simsim).”

{% else %}

Output **nothing at all**. Leave the response blank. Do not write "Nothing", "N/A", or any explanation.

{% endif %}

🔑 Keyword Extraction

Instructions to create a query for keyword/hybrid BM25 search. Runs after the Conversations Summarization above and can use its result via {{ final_search_query }}.

✏ Embeddings Model

Dense Embeddings Weightage

Weightage for dense vs sparse embeddings. 0 for sparse, 1 for dense, 0.5 for equal weight.
Generally speaking, dense embeddings excel at understanding the context of the query, whereas sparse vectors excel at keyword matches.

Max Citations

The maximum number of document search citations.

Max Snippet Words

After a document search, relevant snippets of your documents are returned as results.
This setting adjusts the maximum number of words in each snippet (tokens = words * 2).
A high snippet size allows the LLM to access more information from your document results, at the cost of being verbose and potentially exhausting input tokens (which can cause a failure of the copilot to respond).

Snippet Overlap Ratio

Your knowledge base documents are split into overlapping snippets.
This settings adjusts how much those snippets overlap (overlap tokens = snippet tokens / overlap ratio).
In general you shouldn't need to adjust this.

🔠 Language Model Settings

Avoid Repetition

Response Format

Max Output Tokens

Creativity (aka Sampling Temperature)

Answer Outputs

Run cost = 4 credits

Breakdown: 1 (GPT-4o) + 3/run

With each run, you agree to Gooey.AI's terms & privacy policy.

🐞 Debug

How to Use This Recipe

🙋🏽‍♀️ Need more help? Join our Discord

rmoe's PROD-FarmerAI-MW-GD-CHAT-EN

Instructions

Language Model

Knowledge Base

💪 Capabilities

🦻 Speech Recognition & Translation

Speech-to-Text Provider

Spoken Language

🗣️ Text to Speech & Lipsync

Text-to-Speech Provider

Azure TTS Voice name

🩻 Photo & Document Intelligence

📊 Analytics & Evaluation

🛠️ Developer Tools and Functions

Functions

Variables

Microsoft Server Speech Text to Speech Voice (en-TZ, ImaniNeural):

📄 Knowledge Base

👩‍🏫 Search Instructions

Citation Style

Cache

Create Synthetic Data

Advanced Settings

👁‍🗨 Conversation Summarization

🔑 Keyword Extraction

✏ Embeddings Model

Dense Embeddings Weightage

Max Citations

Max Snippet Words

Snippet Overlap Ratio

🔠 Language Model Settings

Response Format

Max Output Tokens

Creativity (aka Sampling Temperature)

Answer Outputs

How to Use This Recipe

Related Workflows

Lipsync with Text-to-Speech

RAG in the Cloud: Search any document with AI

AI Animation Generator

Compare AI Image Generators

GET STARTED

LEARN

DEVELOPERS

SOCIAL

CONNECT

EXTRAS