FR EN

Personalization

If you want to add search functions to BDS1, it is best to use an LLM that can handle sufficient context, the maximum context displayed by providers is much higher than the context that LLM can handle effectively, most LLM that display a maximum context of 164k do not perform very well when the context exceeds 60k or 80k tokens.

The basic version of BDS1 is designed to work with LLMs such as Deepseek V3.1 Terminus, Minimax M2, or GLM 4.6. For these LLMs, I recommend not adding more than one search function.

The more the search progresses, the more the LLM’s context fills up, so it is better to have 5 search functions in cycles 1, 4 functions in cycles 2 and 3, and only 3 functions in cycle 4, so that the LLM is not overloaded at the end.

To add a function, you must install the tool, select it in the BDS1 agent settings, and add the function command followed by the parameters in parentheses and curly brackets ({}).

If you have vectorized (embedded) documents to create a RAG, to use it you need a tool that searches your RAG and then, for each cycle, adds the command to execute the RAG tool function below the search functions already present in steps 3, 10, 17, and 25.

For example, if the function is called “find”, for the first cycle, you will have the three functions to search the internet followed by the “find” function to search the RAG, which will search based on the keywords A1. Do not provide it with the “query_2” parameter which is specific to the Brained Search tool.

main_xng_smart_search({“query”: “keywords-A1”, “query_2”: “tool-query-A”}) xng_smart_search({“query”: “keywords-A3”, “query_2”: “tool-query-A”}) xng_smart_search({“query”: “keywords-A4”, “query_2”: “tool-query-A”}) find({“query”: “keywords-A1”})

LangSearch provides two free APIs, one for searching like SearXNG or DDGS, and one for reranking (vectorization) that only provides information related to keywords. You can easily add a LangSearch search function to BDS1, tool function is search_and_rerank.

main_xng_smart_search({“query”: “keywords-A1”, “query_2”: “tool-query-A”}) search_and_rerank({“query”: “keywords-A1”}) xng_smart_search({“query”: “keywords-A3”, “query_2”: “tool-query-A”}) xng_smart_search({“query”: “keywords-A4”, “query_2”: “tool-query-A”})