Execute a task with automatic agent discovery and team assembly

Execute a task with automatic agent discovery and team assembly

Usage

genesis(
  task,
  skill_paths = "auto",
  model = "anthropic:claude-3-5-sonnet-20241022",
  cache = TRUE,
  verbose = FALSE,
  architect_model = NULL,
  max_steps = 10,
  mode = c("plan", "direct"),
  use_computer_tools = FALSE
)

Arguments

task

Character string describing the task to accomplish

skill_paths

Character vector of paths to scan for skills, or "auto" for default locations

model

Model to use for agents (default: claude-3-5-sonnet-20241022)

cache

Logical, whether to cache team composition for similar tasks (default: TRUE)

verbose

Logical, whether to print orchestration details (default: FALSE)

architect_model

Model to use for Architect agent (default: same as model)

max_steps

Maximum tool execution steps (default: 10)

mode

Execution mode: "plan" (structured plan-script-execute) or "direct" (single agent)

use_computer_tools

Logical, whether to use computer abstraction layer (default: FALSE). When TRUE, uses atomic tools (bash, read_file, write_file, execute_r_code) instead of loading all skill tools into context. This reduces context window usage by 30-50%.

Value

The result from the team execution

Examples

if (FALSE) { # \dontrun{
# Simple one-line execution
result <- genesis("Analyze the iris dataset and create a scatter plot")

# With computer tools (reduced context usage)
result <- genesis(
  "Analyze the iris dataset",
  use_computer_tools = TRUE
)

# With custom skill paths
result <- genesis(
  "Perform differential expression analysis",
  skill_paths = "~/bioinformatics/skills"
)

# With verbose output to see orchestration
result <- genesis(
  "Check if my package passes CRAN checks",
  verbose = TRUE
)
} # }