{"id":126920,"url":"https://github.com/masstensor/awesome-agent-collaborate-tools","name":"awesome-agent-collaborate-tools","description":"A curated list of tools for AI agent-to-agent communication, coordination, and collaboration","projects_count":86,"last_synced_at":"2026-08-01T22:00:20.272Z","repository":{"id":348656194,"uuid":"1199202818","full_name":"masstensor/awesome-agent-collaborate-tools","owner":"masstensor","description":"A curated list of tools for AI agent-to-agent communication, coordination, and collaboration","archived":false,"fork":false,"pushed_at":"2026-04-02T12:06:35.000Z","size":13,"stargazers_count":1,"open_issues_count":1,"forks_count":2,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-07-13T19:03:25.486Z","etag":null,"topics":["a2a","agent-communication","ai-agents","awesome","awesome-list","llm","mcp","multi-agent","orchestration"],"latest_commit_sha":null,"homepage":"https://github.com/masstensor/awesome-agent-collaborate-tools","language":null,"has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/masstensor.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-04-02T06:12:24.000Z","updated_at":"2026-05-05T08:51:30.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/masstensor/awesome-agent-collaborate-tools","commit_stats":null,"previous_names":["masstensor/awesome-agent-collaborate-tools"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/masstensor/awesome-agent-collaborate-tools","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/masstensor%2Fawesome-agent-collaborate-tools","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/masstensor%2Fawesome-agent-collaborate-tools/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/masstensor%2Fawesome-agent-collaborate-tools/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/masstensor%2Fawesome-agent-collaborate-tools/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/masstensor","download_url":"https://codeload.github.com/masstensor/awesome-agent-collaborate-tools/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/masstensor%2Fawesome-agent-collaborate-tools/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":36171642,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-07-20T02:08:10.276Z","status":"online","status_checked_at":"2026-08-01T02:00:05.789Z","response_time":100,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"created_at":"2026-04-04T18:37:56.777Z","updated_at":"2026-08-01T22:00:20.272Z","primary_language":null,"list_of_lists":false,"displayable":true,"categories":["Identity \u0026 Trust","Multi-Agent Orchestration","State \u0026 Memory","Observability \u0026 Debugging","Developer Tools \u0026 SDKs","Examples \u0026 Demos","Communication Protocols","Messaging \u0026 Notification","Learning Resources","Contributing"],"sub_categories":[],"readme":"# Awesome Agent Collaboration Tools [![Awesome](https://awesome.re/badge.svg)](https://awesome.re)\n\n\u003e A curated list of tools and infrastructure for AI agent-to-agent communication, coordination, and collaboration.\n\nUnlike general \"awesome agents\" lists that focus on *what agents can do*, this list focuses on **how agents work together** — the communication protocols, messaging systems, orchestration frameworks, shared memory layers, identity/trust mechanisms, and observability tools that make multi-agent systems possible.\n\n**2026 context:** MCP has 97M+ downloads, A2A is adopted by 150+ organizations, and multi-agent systems are entering production at scale. This list tracks the infrastructure layer that makes it all work.\n\n## Contents\n\n- [Communication Protocols](#communication-protocols)\n- [Messaging \u0026 Notification](#messaging--notification)\n- [Multi-Agent Orchestration](#multi-agent-orchestration)\n- [State \u0026 Memory](#state--memory)\n- [Identity \u0026 Trust](#identity--trust)\n- [Observability \u0026 Debugging](#observability--debugging)\n- [Developer Tools \u0026 SDKs](#developer-tools--sdks)\n- [Examples \u0026 Demos](#examples--demos)\n- [Learning Resources](#learning-resources)\n\n---\n\n## Communication Protocols\n\nStandards and protocols that define how agents discover, communicate, and delegate tasks to each other.\n\n- [MCP (Model Context Protocol)](https://modelcontextprotocol.io) - Open standard by Anthropic for connecting AI models to tools and data sources. Governed by Linux Foundation. 97M+ downloads. The de facto standard for agent-to-tool communication.\n- [A2A (Agent-to-Agent Protocol)](https://a2a-protocol.org) - Open protocol by Google (donated to Linux Foundation, June 2025) for peer-to-peer agent coordination. 150+ supporting organizations. Peer-to-peer agent communication and task delegation via Agent Cards.\n- [ACP (Agent Communication Protocol)](https://agentcommunicationprotocol.dev) - RESTful agent interoperability protocol originated by IBM/BeeAI, donated to Linux Foundation. Enables agents built with different frameworks (LangChain, CrewAI, AutoGen, etc.) to communicate via a standardized API.\n- [ANP (Agent Network Protocol)](https://agent-network-protocol.com) - Decentralized protocol for open agent networks, built on W3C DID standards. Enables agents to discover and interact without centralized coordination.\n- [UTCP (Universal Tool Calling Protocol)](https://github.com/universal-tool-calling-protocol/utcp-agent) - Exposes tools to models using the tool's native endpoint rather than requiring an MCP wrapper. Contributed to Linux Foundation alongside ACP.\n- [OpenAI Agents SDK — Handoffs](https://platform.openai.com/docs/guides/agents) - First-class agent-to-agent handoff primitives in OpenAI's Agents SDK. Allows agents to delegate tasks with full context transfer.\n\n---\n\n## Messaging \u0026 Notification\n\nReal-time messaging, pub/sub, event streaming, and notification infrastructure designed for or applicable to agent-to-agent communication.\n\n- [IM for Agents](https://im.fengdeagents.site) ⭐ - Agent-to-agent messaging without MCP. Three HTTP calls and your agents are talking — no SDK, no protocol implementation, no infrastructure setup. Cross-framework: Claude, GPT, Gemini, and local LLMs can all join the same room via REST API. Free tier available.\n- [im-agents-mcp](https://github.com/masstensor/im-agents-mcp) - MCP server for IM for Agents. Allows any MCP-compatible agent to send and receive messages from other agents via IM for Agents.\n- [Agent Notify Action](https://github.com/masstensor/agent-notify-action) - GitHub Action for sending notifications from automated agents to humans or other systems. Supports Telegram, Slack, and webhooks.\n- [NATS](https://nats.io) - High-performance, cloud-native messaging system. Subjects and consumer groups make it excellent for agent event streaming and task routing. Supports request-reply patterns common in agent architectures.\n- [Redis Pub/Sub + Streams](https://redis.io/docs/manual/pubsub/) - Battle-tested message patterns. Redis Streams provide persistent, consumer-group-based messaging ideal for agent pipelines where replay and backpressure matter.\n- [RabbitMQ](https://www.rabbitmq.com) - Mature message broker with flexible routing (topics, fanout, headers). Useful for agent task queues with acknowledgment and retry semantics.\n- [Apache Kafka](https://kafka.apache.org) - Distributed event streaming platform. Best for high-throughput, durable agent event logs and audit trails in large-scale multi-agent deployments.\n- [XMTP](https://xmtp.org) - Decentralized, end-to-end encrypted messaging protocol. Used for agent-to-agent messaging with cryptographic identity, especially in Web3 agent contexts.\n\n---\n\n## Multi-Agent Orchestration\n\nFrameworks for defining agent roles, task delegation, coordination logic, and workflow execution across multiple agents.\n\n- [CrewAI](https://github.com/crewAIInc/crewAI) - Role-based multi-agent framework. Agents have roles, goals, and backstories. Simple API, production-ready, 30K+ GitHub stars.\n- [LangGraph](https://github.com/langchain-ai/langgraph) - Graph-based agent orchestration by LangChain. Fine-grained control over agent state machines. Best for complex branching and human-in-the-loop flows.\n- [AutoGen](https://github.com/microsoft/autogen) - Microsoft's framework for conversational multi-agent workflows. Strong support for human-in-the-loop, code generation, and group chat patterns.\n- [Google ADK (Agent Development Kit)](https://google.github.io/adk-docs/) - Google's official SDK for building agents that work with Gemini and A2A protocol. Includes built-in support for multi-agent delegation.\n- [Anthropic Agent SDK](https://docs.anthropic.com/en/docs/agents) - Official Anthropic toolkit for building Claude-powered agents with tool use, multi-turn conversations, and agent handoffs.\n- [OpenAI Agents SDK](https://platform.openai.com/docs/guides/agents) - OpenAI's production agent framework (evolved from Swarm). Built-in support for tool use, handoffs, guardrails, and tracing.\n- [Semantic Kernel](https://github.com/microsoft/semantic-kernel) - Microsoft's open-source SDK for building AI agents and copilots. Supports agent-to-agent patterns via the Process Framework.\n- [Swarms](https://github.com/kyegomez/swarms) - Production-grade swarm orchestration. Supports hierarchical, sequential, concurrent, and mixture-of-agents patterns. Battle-tested for enterprise deployments.\n- [AgentScope](https://github.com/modelscope/agentscope) - Alibaba's multi-agent platform. Provides MsgHub and pipeline abstractions for efficient message routing between agents.\n- [Mastra](https://github.com/mastra-ai/mastra) - TypeScript-first agent framework with memory, tool-calling, workflows, and RAG. Designed for building production AI applications with multi-agent capabilities.\n- [smolagents](https://github.com/huggingface/smolagents) - Hugging Face's minimal, code-centric agent framework. Agents write and execute Python code to achieve goals. Simple integration into multi-agent pipelines.\n- [open-multi-agent](https://github.com/JackChen-me/open-multi-agent) - Production-grade multi-agent orchestration. Model-agnostic, supports team collaboration, task scheduling, and inter-agent communication with shared memory.\n- [PydanticAI](https://github.com/pydantic/pydantic-ai) - Production-grade agent framework from the Pydantic team. Type-safe tool calling, dependency injection, and structured output. Supports multi-agent orchestration via agent handoff primitives.\n- [LlamaIndex Workflows](https://docs.llamaindex.ai/en/stable/module_guides/workflow/) - Event-driven, async multi-agent orchestration built into LlamaIndex. Define agents as workflow steps with typed events — clean model for complex agent DAGs.\n- [Dify](https://github.com/langgenius/dify) - Open-source LLM app platform with visual agent builder. Supports multi-agent workflows, tool integration, and knowledge retrieval. 60K+ GitHub stars.\n\n---\n\n## State \u0026 Memory\n\nTools for persisting agent context, shared knowledge, cross-session memory, and distributed state management.\n\n- [Mem0](https://github.com/mem0ai/mem0) - Adaptive memory layer for AI agents. Intelligently compresses history into optimized memory representations. Claims up to 80% prompt token reduction. Lowest integration friction for standalone memory.\n- [Zep](https://github.com/getzep/zep) - Long-term memory store with temporal knowledge graph. Tracks how facts change over time. Integrates structured business data with conversational history. Best for temporal-aware production pipelines.\n- [Letta (formerly MemGPT)](https://github.com/cpacker/MemGPT) - Agents with OS-style virtual memory management. Self-editing memory architecture with in-context vs archival storage. Includes REST API and development environment.\n- [Cognee](https://github.com/topoteretes/cognee) - Knowledge-graph-first memory for agents. Builds structured knowledge graphs from unstructured data for rich semantic retrieval.\n- [LangMem](https://github.com/langchain-ai/langmem) - LangChain's memory abstraction layer. Multiple memory types (episodic, semantic, procedural) with a unified API.\n- [Supermemory](https://github.com/supermemoryai/supermemory) - Universal memory API for AI agents. Import from anywhere, query with semantic search. Designed as a \"memory backbone\" across multiple agents.\n- [Redis (State Management)](https://redis.io) - Fast in-memory store for short-term agent state, session data, and distributed locks. Essential for stateful multi-agent systems requiring sub-millisecond access.\n- [Chroma](https://github.com/chroma-core/chroma) - Open-source embedding database. The most popular vector store for agent memory. Simple API, runs locally or in the cloud. pip install chromadb.\n- [Qdrant](https://github.com/qdrant/qdrant) - High-performance vector database written in Rust. Excellent for production agent long-term memory with filtering, payload indexing, and sparse vector support.\n- [Weaviate](https://github.com/weaviate/weaviate) - Open-source vector database with built-in vectorization. Modules for text, images, and multi-modal memory. Native GraphQL API.\n- [Pinecone](https://www.pinecone.io) - Managed vector database. Serverless option allows agents to query semantic memory with zero infrastructure. Widely used in production agent deployments.\n- [Haystack DocumentStore](https://docs.haystack.deepset.ai/docs/document-store) - Pluggable document store interface for Haystack agents. Supports Elasticsearch, OpenSearch, Weaviate, Pinecone, and more. Uniform API across backends.\n\n---\n\n## Identity \u0026 Trust\n\nSpecifications and tools for establishing agent identity, authentication, and trust relationships between agents. The defining security challenge of 2026.\n\n- [A2A Agent Cards](https://a2a-protocol.org/latest/topics/agent-discovery/) - A2A protocol standard for agent self-description. Agents publish capabilities, authentication requirements, and endpoints in a standardized JSON format. Enables dynamic agent discovery.\n- [SPIFFE / SPIRE](https://spiffe.io) - Production-grade workload identity framework. Each agent gets a cryptographically verifiable SVID (SPIFFE Verifiable Identity Document) — no shared secrets. SPIRE handles issuance, rotation, and revocation. The current best practice for secure agent-to-agent mTLS.\n- [WIMSE (Workload Identity and Minimal Secrets)](https://datatracker.ietf.org/doc/draft-ietf-wimse-arch/) - IETF working group standard for workload/agent identity in cloud environments. Combines SPIFFE, OAuth 2.0, and token exchange for agent delegation. First draft published 2025, actively developed 2026.\n- [OAuth 2.0 for AI Agents](https://oauth.net/2/) - Using OAuth 2.0 / DPoP for agent-to-agent authorization. Best practice for agents calling APIs or other agents on behalf of users. IETF draft (draft-klrc-aiagent-auth-00) published March 2026.\n- [DID (Decentralized Identifiers)](https://www.w3.org/TR/did-core/) - W3C standard for cryptographic agent identity without central authority. Used as the foundation for ANP and decentralized agent networks.\n- [WorkOS](https://workos.com) - Enterprise identity and auth platform increasingly used for AI agent authorization. Fine-grained permissions, audit logs, and organization-level access control — applies cleanly to multi-agent systems where agents act on behalf of users.\n- [Permit.io](https://permit.io) - Policy-based authorization for apps and agents. Define what agents are allowed to do via ReBAC/ABAC policies. Supports runtime policy updates without redeploy.\n\n---\n\n## Observability \u0026 Debugging\n\nTools for tracing, monitoring, evaluating, and debugging multi-agent systems in production.\n\n- [LangSmith](https://smith.langchain.com) - End-to-end observability platform for LLM/agent applications by LangChain. Tracing, evaluation, dataset management, and prompt experimentation.\n- [Langfuse](https://github.com/langfuse/langfuse) - Open-source LLM observability (MIT license). Acquired by ClickHouse. 19 Fortune 50 clients. Full-featured: tracing, evaluations, prompt management, datasets.\n- [Arize Phoenix](https://github.com/Arize-ai/phoenix) - Open-source LLM observability. Trace agent runs, visualize spans, and evaluate outputs locally or in the cloud. Built on OpenTelemetry.\n- [AgentOps](https://github.com/AgentOps-AI/agentops) - Agent-specific observability and testing. Session replays, cost tracking, failure analysis, and compliance monitoring. Native integrations with CrewAI, AutoGen, LangChain.\n- [Maxim AI](https://www.getmaxim.ai) - Agent quality and observability platform. Evaluation pipelines, red-teaming, and production monitoring for agentic workflows.\n- [OpenTelemetry](https://opentelemetry.io) - Vendor-neutral observability framework. Use with [OpenLLMetry](https://github.com/traceloop/openllmetry) for LLM-specific spans and agent traces.\n- [Weights \u0026 Biases Weave](https://weave-docs.wandb.ai) - W\u0026B's LLM observability platform. Trace, evaluate, and monitor agent runs with rich visualization.\n- [Traceloop / OpenLLMetry](https://github.com/traceloop/openllmetry) - OpenTelemetry-based LLM observability. Automatic instrumentation for LangChain, CrewAI, AutoGen, and 20+ frameworks. Sends traces to any OTel-compatible backend (Datadog, Grafana, etc.).\n- [Helicone](https://github.com/Helicone/helicone) - Open-source LLM observability and caching proxy. Drop-in for OpenAI/Anthropic/Gemini. Per-request latency, cost, and error tracking with user-level attribution.\n- [Honeyhive](https://honeyhive.ai) - Agent evaluation and monitoring platform. Dataset management, A/B testing for prompts, and production quality monitoring. Focus on multi-step agent workflows.\n\n---\n\n## Developer Tools \u0026 SDKs\n\nSDKs, CLIs, and utilities that accelerate building multi-agent systems.\n\n- [im-for-agents-python](https://github.com/masstensor/im-for-agents-python) - Python SDK for IM for Agents. Send and receive messages between agents with minimal boilerplate.\n- [multi-agent-starter](https://github.com/masstensor/multi-agent-starter) - Starter template for building multi-agent applications with MCP, A2A, and IM for Agents integration out of the box.\n- [Dapr](https://dapr.io) - Distributed Application Runtime. Service invocation, pub/sub, state management, and actor pattern — all as portable building blocks for agent microservices.\n- [BeeAI Framework](https://github.com/i-am-bee/bee-agent-framework) - IBM's open-source agent framework powering the ACP protocol. Production-ready with tool use, memory, and structured output.\n- [FastMCP](https://github.com/jlowin/fastmcp) - The fastest way to build MCP servers in Python. Decorator-based API similar to FastAPI. Used by tens of thousands of MCP server authors. Now part of the official MCP Python SDK.\n- [Composio](https://github.com/composiohq/composio) - Managed tool integration platform for AI agents. 250+ pre-built integrations (GitHub, Slack, Notion, etc.) with auth management. Works with LangChain, CrewAI, AutoGen, and any agent framework.\n- [Portkey AI Gateway](https://github.com/Portkey-AI/gateway) - Open-source AI gateway with fallbacks, retries, load balancing, and request routing across 200+ LLMs. Essential for multi-agent deployments that use multiple model providers.\n- [LiteLLM](https://github.com/BerriAI/litellm) - Unified API for 100+ LLMs. Use the same code to call OpenAI, Anthropic, Gemini, Groq, Ollama. Simplifies multi-agent systems where different agents use different models.\n- [instructor](https://github.com/jxnl/instructor) - Structured output extraction from LLMs using Pydantic. Critical for agent-to-agent communication with typed message schemas — ensures agents receive parseable, validated data.\n- [Arcade AI](https://github.com/ArcadeAI/arcade-ai) - Managed tool execution platform for AI agents. Secure, sandboxed execution with auth flows for enterprise tools. Reduces the surface area for agent security incidents.\n\n---\n\n## Examples \u0026 Demos\n\nReference implementations and worked examples of multi-agent collaboration patterns.\n\n- [a2a-examples](https://github.com/masstensor/a2a-examples) - Practical examples of A2A protocol agent communication. Covers agent discovery, task delegation, and streaming responses.\n- [multi-agent-demo](https://github.com/masstensor/multi-agent-demo) - End-to-end demo of multiple agents collaborating via IM for Agents and MCP. Shows orchestrator + specialist agent pattern.\n- [multi-agent-patterns](https://github.com/masstensor/multi-agent-patterns) - Common patterns for building multi-agent systems with code examples. Covers pipeline, fan-out, consensus, and critic patterns.\n- [Google A2A Samples](https://github.com/google-a2a/a2a-samples) - Official Google samples for A2A protocol. Best starting point for understanding Agent Cards and task lifecycle.\n- [AutoGen Examples](https://github.com/microsoft/autogen/tree/main/samples) - Microsoft's reference examples for AutoGen agent patterns including group chat and code execution.\n- [LangGraph Multi-Agent Examples](https://github.com/langchain-ai/langgraph/tree/main/examples) - Official LangGraph examples covering supervisor, hierarchical, and collaborative multi-agent patterns with state management.\n- [CrewAI Examples](https://github.com/crewAIInc/crewAI-examples) - Official CrewAI examples including trip planner, stock analysis, and game builder crews. Best starting point for role-based multi-agent systems.\n- [IM for Agents — 3 curl demo](https://masstensor.github.io/im-for-agents/) - Minimal demo of agent-to-agent messaging with 3 HTTP calls — no SDK, no framework. Shows the baseline for any cross-framework agent coordination.\n\n---\n\n## Learning Resources\n\nArticles, specs, and guides for understanding the multi-agent collaboration landscape.\n\n- [A2A Protocol Specification](https://a2a-protocol.org/latest/) - Full A2A specification, concepts, and implementation guide.\n- [MCP Documentation](https://modelcontextprotocol.io/docs) - Official MCP docs with quickstarts, tutorials, and server implementation guides.\n- [MCP vs A2A: Complete Guide (2026)](https://dev.to/pockit_tools/mcp-vs-a2a-the-complete-guide-to-ai-agent-protocols-in-2026-30li) - Deep comparison of when to use MCP vs A2A vs ACP.\n- [AI Agent Protocol Ecosystem Map 2026](https://www.digitalapplied.com/blog/ai-agent-protocol-ecosystem-map-2026-mcp-a2a-acp-ucp) - Visual overview of the complete agent protocol landscape.\n- [Survey of Agent Interoperability Protocols](https://arxiv.org/html/2505.02279v1) - Academic survey covering MCP, ACP, A2A, and ANP.\n- [120+ Agentic AI Tools Mapped (2026)](https://www.stackone.com/blog/ai-agent-tools-landscape-2026/) - Comprehensive landscape map across 11 agent tool categories.\n- [4 Patterns AI Agents Use to Coordinate](https://masstensor.github.io/im-for-agents/blog-coordination-patterns.html) - Practical guide: shared files, MCP, Kafka, and REST messaging — when each pattern breaks in production.\n- [The Agent Identity Playbook (2026)](https://www.strata.io/blog/agentic-identity/new-identity-playbook-ai-agents-not-nhi-8b/) - How to think about non-human identity (NHI) for AI agents in enterprise environments.\n- [Securing AI Agents: The Defining Cybersecurity Challenge of 2026](https://www.bvp.com/atlas/securing-ai-agents-the-defining-cybersecurity-challenge-of-2026) - Bessemer VP analysis of the agent security landscape — permissions, secrets, and trust boundaries.\n\n---\n\n## Contributing\n\nSubmissions welcome! Please read the [contribution guidelines](CONTRIBUTING.md) first.\n\nOnly add tools that are:\n- Actively maintained (last commit within 12 months, or commercially supported)\n- Genuinely useful for agent-to-agent collaboration (not just \"agent-adjacent\")\n- Open source, or have a meaningful free tier\n\n---\n\n*Maintained by [@masstensor](https://github.com/masstensor) · [IM for Agents](https://im.fengdeagents.site) — messaging infrastructure for AI agents*\n","projects_url":"https://awesome.ecosyste.ms/api/v1/lists/masstensor%2Fawesome-agent-collaborate-tools/projects"}