dragonfarm's tools for an ellmer chat or an MCP client — dragon_ellmer

Turns dragonfarm's tool contracts (search a document set, answer from it, read documents, list sets and runs, check a run) into ellmer tools, so any ellmer chat, with any model, can use your documents and runs. Every call goes through the same checks as inside dragonfarm: arguments are checked against the tool's contract, an agent is held to the current policy (some calls are refused, some wait for a person), and every call is recorded in the catalogue's event log.

Usage

dragon_ellmer_tools(agent = NULL, tools = NULL)

Arguments

agent: Name the caller as an agent in the policy (for example "answer"), so its calls are checked against what that agent may do. NULL (the default) means a person acting directly.
tools: Which tools to include, by name. NULL includes all.

Value

A list of ellmer tool definitions.

Details

To offer them to Claude Desktop, Claude Code, or another MCP client, serve them with the mcptools package, for example mcptools::mcp_server(tools = dragon_ellmer_tools()).

Examples

if (requireNamespace("ellmer", quietly = TRUE)) {
  names(dragon_ellmer_tools())
}
#>  [1] "answer_question"          "build_training_set"      
#>  [3] "cancel_run"               "check_answer"            
#>  [5] "check_licence"            "deletion_impact"         
#>  [7] "evaluate_run"             "list_document_sets"      
#>  [9] "list_runs"                "promote_model"           
#> [11] "provenance_report"        "read_documents"          
#> [13] "recall_training"          "run_status"              
#> [15] "scan_personal_data"       "search_documents"        
#> [17] "train_model"              "verify_document_set"     
#> [19] "write_practice_questions"
if (FALSE) { # \dontrun{
chat <- ellmer::chat_anthropic()
chat$register_tools(dragon_ellmer_tools())
chat$chat("What do my FDA guidances say about covariate adjustment?")
} # }