govkit-multi-agent-designlisted

# Governed Agent Topology Design Determine the feature name from the user's request; if it is not provided, ask before proceeding. ## Required extension Confirm `extensions/skill-oriented-agent-architecture/manifest.yaml` exists. If it is missing, stop and report: ```bash govkit extension add skill-oriented-agent-architecture --target . ``` ## Inputs to read - `features/<feature_name>/nfrs.md` - `features/<feature_name>/acceptance.feature` - `features/<feature_name>/eval_criteria.yaml` - `extensions/skill-oriented-agent-architecture/docs/backend/architecture/SKILL_ORIENTED_AGENT_ARCHITECTURE.md` - `extensions/skill-oriented-agent-architecture/docs/backend/architecture/RUNTIME_STATE_AND_EXECUTION_CONTRACT.md` - `extensions/skill-oriented-agent-architecture/docs/backend/architecture/AUTHORITY_AND_APPROVAL_CONTRACT.md` - `extensions/skill-oriented-agent-architecture/docs/backend/architecture/RESILIENCE_AND_RECOVERY_CONTRACT.md` - `extensions/skill-oriented-agent-architecture/docs/backend/architecture/EVALUATION_EVIDENCE_AND_COMPLETION_CONTRACT.md` - `docs/backend/architecture/TECH_STACK.md` and applicable ADRs If the design invokes language models, also load the applicable `llm-application` contracts. ## 1. Justify the topology State why multiple agent runs, delegation, or specialization is necessary. Prefer the smallest topology that satisfies the feature. Stop and recommend a simpler workflow when separate agent ownership does not provide a clear benefit. ## 2. Defin