pub struct AgentConfig {Show 13 fields
pub id: String,
pub enabled: SwitchPosition,
pub display_name: Option<String>,
pub description: Option<String>,
pub tools: Vec<String>,
pub skills: Vec<String>,
pub prompt_path: Path<'static>,
pub prompt_markdown: Option<String>,
pub chat_config: ChatConfig,
pub user_instructions_enabled: bool,
pub agents_md_enabled: bool,
pub agents_md_path: Option<String>,
pub user_instructions_hook: Option<String>,
}Expand description
Configuration for a single agent declared in agents.conf.
Each agent has an identity, an enablement switch, a system prompt, a list of tools it may invoke, and tuning knobs for user-instruction injection and extended-thinking budgets.
Fields§
§id: StringUnique identifier for this agent (the rc.conf service name).
enabled: SwitchPositionWhether the agent is enabled, disabled, or requires manual confirmation.
display_name: Option<String>Human-readable display name, if specified.
description: Option<String>Short prose description of the agent’s purpose.
tools: Vec<String>Tool identifiers this agent is allowed to invoke.
skills: Vec<String>Skill identifiers this agent has access to.
prompt_path: Path<'static>Filesystem path to the agent’s system prompt markdown file.
prompt_markdown: Option<String>Loaded system prompt markdown content, or None when the file is absent.
chat_config: ChatConfigMerged chat configuration (model, thinking budget, etc.).
user_instructions_enabled: boolWhether user-instruction injection is enabled for this agent.
agents_md_enabled: boolWhether the AGENTS.md file should be appended to the system prompt.
agents_md_path: Option<String>Explicit override path for the AGENTS.md file, if set.
user_instructions_hook: Option<String>Shell command executed as a hook to produce additional user instructions.