Index

skillorchestra

SkillOrchestra — faithful port of the eval-orchestrator (arXiv:2602.19672).

Importing this package registers the skillorchestra agent. Layout mirrors the upstream repo (external/SkillOrchestra):

• :mod:.prompts — verbatim eval_orchestrator / model_routing / learning prompt templates.
• :mod:.stage_router — verbatim StageSkillHandbook, parse_skill_analysis, the 5 routing strategies.
• :mod:.types — verbatim learning-time data types (BetaCompetence, Skill, AgentProfile, ...).
• :mod:.pool — model-alias -> local/cloud resolution.
• :mod:.tools — search / enhance_reasoning / answer executors.
• :mod:.orchestrator — the multi-round search->code->answer loop.
• :mod:.agent — :class:SkillOrchestraAgent, the harness entry.

Classes

SkillOrchestraAgent

SkillOrchestraAgent(engine: InferenceEngine, model: str, *, local_model: Optional[str] = None, local_endpoint: Optional[str] = None, cloud_endpoint: str = 'anthropic', cfg: Optional[Dict[str, Any]] = None, bus: Optional[Any] = None, temperature: Optional[float] = None, max_tokens: Optional[int] = None)

Bases: LocalCloudAgent

Inference-time skill-aware orchestrator. See module docstring.

Source code in src/openjarvis/agents/hybrid/_base.py

def __init__(
    self,
    engine: InferenceEngine,
    model: str,
    *,
    local_model: Optional[str] = None,
    local_endpoint: Optional[str] = None,
    cloud_endpoint: str = "anthropic",
    cfg: Optional[Dict[str, Any]] = None,
    bus: Optional[Any] = None,
    temperature: Optional[float] = None,
    max_tokens: Optional[int] = None,
) -> None:
    super().__init__(
        engine,
        model,
        bus=bus,
        temperature=temperature,
        max_tokens=max_tokens,
    )
    self._cloud_model = model
    self._cloud_endpoint = (cloud_endpoint or "anthropic").lower()
    self._local_model = local_model
    self._local_endpoint = local_endpoint
    self._cfg: Dict[str, Any] = dict(cfg or {})