Skip to content

techniques

techniques

Technique system configuration models.

Techniques are composable components attached to agent entities — an ECS pattern for AI agents. Each technique is independently reusable across projects, agents, and scores.

Three kinds: - skill: Text-based methodology (memory protocol, mateship, coordination) - mcp: MCP server tools accessible via shared pool (github, filesystem) - protocol: Communication protocols (A2A)

Each technique declares which phases of the agent cycle it's available in, allowing the compiler to inject technique manifests per-phase.

Classes

TechniqueKind

Bases: str, Enum

Kind of technique component.

Maps to the ECS component taxonomy: - SKILL: Text-based methodology injected as cadenza context - MCP: MCP server tools accessible via the shared pool - PROTOCOL: Communication protocols (A2A, coordination)

TechniqueConfig

Bases: BaseModel

Configuration for a single technique attached to an agent.

Techniques are composable: an agent can have multiple techniques of different kinds. The compiler's technique wirer reads these declarations and injects the appropriate manifests, MCP access, and protocol config into each phase's cadenza context.

Example YAML::

techniques:
  a2a:
    kind: protocol
    phases: [recon, plan, work, integration, inspect, aar]
  github:
    kind: mcp
    phases: [recon, work, integration]
    config:
      server: github
      transport: stdio