Expand description
Multi-agent orchestration for the TraitClaw AI agent framework.
Provides Team structs for composing agents, Router protocol for
message routing, delegation between agents, and VerificationChain
for generate-then-verify patterns.
§Quick Start
use traitclaw_team::{AgentRole, Team, VerificationChain, VerifyResult};
use traitclaw_team::router::SequentialRouter;
let team = Team::new("research_team")
.add_role(AgentRole::new("researcher", "Research topics in depth"))
.add_role(AgentRole::new("writer", "Write clear summaries"));
assert_eq!(team.name(), "research_team");
assert_eq!(team.roles().len(), 2);Re-exports§
pub use conditional_router::ConditionalRouter;pub use execution::run_verification_chain;pub use execution::TeamRunner;pub use team_context::TeamContext;
Modules§
- conditional_
router ConditionalRouter— content-based routing using regex patterns.- execution
- Team execution engine —
Team.bind()andTeam.run(). - group_
chat - Multi-agent group chat with configurable turn-taking and termination.
- router
- Router — pluggable multi-agent message routing.
- team_
context - Shared
TeamContext— thread-safe inter-agent state store.
Structs§
- Agent
Role - A role within a team — describes what an agent specializes in.
- Team
- A team of agents working together.
- Verification
Chain - A verification chain: generate with one agent, verify with another.
Enums§
- Verify
Result - Result of a verification step.
Functions§
- pool_
from_ team - Create an
AgentPoolfrom aTeamand a provider. - pool_
from_ team_ arc - Create an
AgentPoolfrom aTeamusing a pre-wrappedArc<dyn Provider>.