{"id":"vibeeval-agentica-infrastructure","name":"agentica-infrastructure","summary":"AgenticaマルチエージェントインフラストラクチャAPIの参考ガイド","body":"# Agentica Infrastructure Reference\r\n\r\nComplete API specification for Agentica multi-agent coordination infrastructure.\r\n\r\n## When to Use\r\n\r\n- Building multi-agent workflows with Agentica patterns\r\n- Need exact constructor signatures for pattern classes\r\n- Want to understand coordination database schema\r\n- Implementing custom patterns using primitives\r\n- Debugging agent tracking or orphan detection\r\n\r\n## Quick Reference\r\n\r\n### 11 Pattern Classes\r\n\r\n| Pattern | Purpose | Key Method |\r\n|---------|---------|------------|\r\n| `Swarm` | Parallel perspectives | `.execute(query)` |\r\n| `Pipeline` | Sequential stages | `.run(initial_state)` |\r\n| `Hierarchical` | Coordinator + specialists | `.execute(task)` |\r\n| `Jury` | Voting consensus | `.decide(return_type, question)` |\r\n| `GeneratorCritic` | Iterative refinement | `.run(task)` |\r\n| `CircuitBreaker` | Failure fallback | `.execute(query)` |\r\n| `Adversarial` | Debate + judge | `.resolve(question)` |\r\n| `ChainOfResponsibility` | Route to handler | `.process(query)` |\r\n| `MapReduce` | Fan out + reduce | `.execute(query, chunks)` |\r\n| `Blackboard` | Shared state | `.solve(query)` |\r\n| `EventDriven` | Event bus | `.publish(event)` |\r\n\r\n### Core Infrastructure\r\n\r\n| Component | File | Purpose |\r\n|-----------|------|---------|\r\n| `CoordinationDB` | `coordination.py` | SQLite tracking |\r\n| `tracked_spawn` | `tracked_agent.py` | Agent with tracking |\r\n| `HandoffAtom` | `handoff_atom.py` | Universal handoff format |\r\n| `BlackboardCache` | `blackboard.py` | Hot tier communication |\r\n| `MemoryService` | `memory_service.py` | Core + Archival memory |\r\n| `create_claude_scope` | `claude_scope.py` | Scope with file ops |\r\n\r\n### Primitives\r\n\r\n| Primitive | Purpose |\r\n|-----------|---------|\r\n| `Consensus` | Voting (MAJORITY, UNANIMOUS, THRESHOLD) |\r\n| `Aggregator` | Combine results (MERGE, CONCAT, BEST) |\r\n| `HandoffState` | Structured agent handoff |\r\n| `build_premise` | Structured premise builder |\r\n| `gather_fail_fast` | TaskGroup-based parallel execution |\r\n\r\n## Full API Spec\r\n\r\nSee: `API_SPEC.md` in this skill directory\r\n\r\n## Usage Example\r\n\r\n```python\r\nfrom scripts.agentica_patterns.patterns import Swarm, Jury\r\nfrom scripts.agentica_patterns.primitives import ConsensusMode\r\nfrom scripts.agentica_patterns.coordination import CoordinationDB\r\nfrom scripts.agentica_patterns.tracked_agent import tracked_spawn\r\n\r\n# Create tracking database\r\ndb = CoordinationDB(session_id=\"my-session\")\r\n\r\n# Swarm with tracking\r\nswarm = Swarm(\r\n    perspectives=[\"Security expert\", \"Performance expert\"],\r\n    db=db\r\n)\r\nresult = await swarm.execute(\"Review this code\")\r\n\r\n# Jury with consensus\r\njury = Jury(\r\n    num_jurors=3,\r\n    consensus_mode=ConsensusMode.MAJORITY,\r\n    premise=\"You evaluate code quality\",\r\n    db=db\r\n)\r\nverdict = await jury.decide(bool, \"Is this code production ready?\")\r\n```\r\n\r\n## Location\r\n\r\nAPI spec: `.claude/skills/agentica-infrastructure/API_SPEC.md`\r\nSource: `scripts/agentica_patterns/`","author":"@vibeeval","ownerProfile":null,"authorContacts":null,"sourceUrl":"https://github.com/vibeeval/vibecosystem/tree/main/skills/agentica-infrastructure","license":"MIT","category":"devops","lang":"en","tokens":693,"stars":0,"calls30d":1,"claimed":false,"visibility":"public","origin":"crawler","version":"0.1.0","createdAt":"2026-08-22","updatedAt":"2026-08-22","files":[{"path":"API_SPEC.md","size":27621,"sha256":"026aab4e6567092606611c90ff0d78d444079971a2607d2d9317fc3d5d7da132"}],"requires":{"mcp":[],"tools":["Read"]},"safety":{"flags":[],"scannedAt":"2026-08-22","hasScripts":false,"networkEndpoints":[]}}