6 ms·

This is interesting. Also "Discover tools on-demand". Are there any stats or estimates how many tools an LLM / agent could handle with this approach vs. loading

by Fannon 10mo ago

This is interesting. Also "Discover tools on-demand". Are there any stats or estimates how many tools an LLM / agent could handle with this approach vs. loading them all into context as MCP tools?

thecopy 10mo ago
What i have read its in the range of 60-80. (shameless plug: im building an cloud based gateway where the set of servers given to an mcp client can be controlled using "profiles": https://docs.gatana.ai/profiles/ https://docs.gatana.ai/profiles/)