Agentica Infrastructure

parcadei/continuous-claude-v3/.claude/skills/agentica-infrastructure

作者 parcadeid07ff4b06b62无许可证3.9K 个星标收录于 2026年10月8日更新于 2026年10月8日仓库8个月前更新

Reference guide for Agentica multi-agent infrastructure APIs

仅含说明AI & Agents
AI 生成的概览

记录 Agentica 多智能体基础设施 API、模式类、原语与协调组件的参考指南。

功能
提供 Agentica 多智能体协调基础设施的参考规范,列出 Swarm、Pipeline、Jury 等十一个模式类及其关键方法。它还记录 CoordinationDB、tracked_spawn、HandoffAtom 等核心基础设施组件,以及 Consensus、Aggregator 等原语。完整 API 规范位于随附的 API_SPEC.md 文件中,技能内还包含一个 Python 使用示例。
适用场景
适用于使用 Agentica 模式构建多智能体工作流,或需要模式类精确构造函数签名时。也用于理解协调数据库架构、使用原语实现自定义模式,或调试智能体跟踪与孤立检测。
运行要求
仅为说明文档,不运行脚本。它引用技能目录中的 API_SPEC.md 以及 scripts/agentica_patterns/ 下的源文件,使用示例从该路径导入 Python 模块。

Agentica Infrastructure Reference

Complete API specification for Agentica multi-agent coordination infrastructure.

When to Use

  • Building multi-agent workflows with Agentica patterns
  • Need exact constructor signatures for pattern classes
  • Want to understand coordination database schema
  • Implementing custom patterns using primitives
  • Debugging agent tracking or orphan detection

Quick Reference

11 Pattern Classes

PatternPurposeKey Method
SwarmParallel perspectives.execute(query)
PipelineSequential stages.run(initial_state)
HierarchicalCoordinator + specialists.execute(task)
JuryVoting consensus.decide(return_type, question)
GeneratorCriticIterative refinement.run(task)
CircuitBreakerFailure fallback.execute(query)
AdversarialDebate + judge.resolve(question)
ChainOfResponsibilityRoute to handler.process(query)
MapReduceFan out + reduce.execute(query, chunks)
BlackboardShared state.solve(query)
EventDrivenEvent bus.publish(event)

Core Infrastructure

ComponentFilePurpose
CoordinationDBcoordination.pySQLite tracking
tracked_spawntracked_agent.pyAgent with tracking
HandoffAtomhandoff_atom.pyUniversal handoff format
BlackboardCacheblackboard.pyHot tier communication
MemoryServicememory_service.pyCore + Archival memory
create_claude_scopeclaude_scope.pyScope with file ops

Primitives

PrimitivePurpose
ConsensusVoting (MAJORITY, UNANIMOUS, THRESHOLD)
AggregatorCombine results (MERGE, CONCAT, BEST)
HandoffStateStructured agent handoff
build_premiseStructured premise builder
gather_fail_fastTaskGroup-based parallel execution

Full API Spec

See: API_SPEC.md in this skill directory

Usage Example

python
from scripts.agentica_patterns.patterns import Swarm, Juryfrom scripts.agentica_patterns.primitives import ConsensusModefrom scripts.agentica_patterns.coordination import CoordinationDBfrom scripts.agentica_patterns.tracked_agent import tracked_spawn
# Create tracking databasedb = CoordinationDB(session_id="my-session")
# Swarm with trackingswarm = Swarm(    perspectives=["Security expert", "Performance expert"],    db=db)result = await swarm.execute("Review this code")
# Jury with consensusjury = Jury(    num_jurors=3,    consensus_mode=ConsensusMode.MAJORITY,    premise="You evaluate code quality",    db=db)verdict = await jury.decide(bool, "Is this code production ready?")

Location

API spec: .claude/skills/agentica-infrastructure/API_SPEC.md Source: scripts/agentica_patterns/

来源与署名

来源:parcadei/continuous-claude-v3位于.claude/skills/agentica-infrastructure提交d07ff4b

许可证: 无许可证

内容归原作者所有。SourceWeft 从公开仓库中收录这些内容。

举报或申请下架