Struct AgentConfig 

pub struct AgentConfig {

    pub command: String,
    pub provider: Option<String>,
    pub model: Option<String>,
    pub system_prompt: Option<String>,
    pub root: Option<String>,
    pub add_dirs: Vec<String>,
    pub env: Vec<(String, String)>,
    pub files: Vec<String>,
    pub auto_approve: bool,
    pub worktree: Option<Option<String>>,
    pub sandbox: Option<String>,
    pub json_mode: bool,
    pub json_schema: Option<String>,
    pub output_format: Option<String>,
    pub max_turns: Option<u32>,
    pub timeout: Option<String>,
    pub mcp_config: Option<String>,
    pub session_name: String,
    pub description: Option<String>,
    pub tags: Vec<String>,
    pub prompt: String,
    pub accepts_agent_args: bool,
    pub command_params: Option<CommandParams>,
    pub interactive: bool,
}

Serializable snapshot of everything a step contributes to an agent invocation. Produced by [build_agent_config] and used by both the executor (to configure a AgentBuilder) and --dry-run (to show the author what each step will ask of the agent).

The shape mirrors AgentBuilder field-for-field; additional command-specific parameters (review/plan/pipe/collect/summary) are stored in the optional command_params bag.

Fields

command: String

Subcommand label — "run", "review", "plan", "pipe", "collect", or "summary".

provider: Option<String>

Agent-level knobs.

model: Option<String>

system_prompt: Option<String>

root: Option<String>

add_dirs: Vec<String>

env: Vec<(String, String)>

files: Vec<String>

auto_approve: bool

worktree: Option<Option<String>>

None → no worktree, Some(None) → generated name, Some(Some(n)) → explicit.

sandbox: Option<String>

json_mode: bool

Output shaping.

json_schema: Option<String>

output_format: Option<String>

max_turns: Option<u32>

Turn / timeout / MCP.

timeout: Option<String>

mcp_config: Option<String>

session_name: String

Session metadata — always set (session name derives from workflow/step).

description: Option<String>

tags: Vec<String>

prompt: String

The effective prompt after dependency/context/plan prepending.

accepts_agent_args: bool

Whether accepts_agent_args was true for this command — pipe/run/ review/plan/exec respect agent-level flags; collect/summary don’t.

command_params: Option<CommandParams>

Extra params for the non-plain commands. None for run.

interactive: bool

Interactive flag — steps with interactive = true run through AgentBuilder::run instead of exec.

Trait Implementations

impl Clone for AgentConfig

fn clone(&self) -> AgentConfig

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for AgentConfig

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for AgentConfig

fn default() -> AgentConfig

Returns the “default value” for a type.

impl Serialize for AgentConfig

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.