Category
366 tools
Airtable bases via MCP. Read records, create rows, update fields, manage views. Good for low-code agent workflows.
Sandboxed code execution for agents. Run Python, JS, and shell commands in secure cloud sandboxes. The gold standard for agent code execution safety.
Pre-built agent templates using E2B sandboxes. Code-executing agents with safe isolation.
Manage Docker containers and images via MCP. Start, stop, inspect, and build containers from agent workflows.
200+ integrations, zero-config managed auth. The most production-ready MCP server available — full OpenAPI 3.1 compliance, zero critical failure modes found in testing.
OpenAI's production agent framework. Handoffs, guardrails, streaming, tracing built-in. The most polished commercial framework available.
Allows AI agents to access Figma files, designs, and components through MCP protocol.
A fully managed AWS service for building and deploying autonomous agents with automatic scaling, prompt engineering, memory management, and RAG capabilities.
Google-originated open protocol for agent-to-agent communication. Standardizes task delegation, capability discovery, and result passing between autonomous agents.
Official Linear MCP integration. Create issues, manage projects, query team workflows. Clean API surface with good error handling.
Enables AI agents to query dashboards, metrics, and visualizations from Grafana via MCP.
An SDK for building agents with Vertex AI integration, A2A protocol support, and Gemini model access for enterprise deployment.
Connects AI agents to Datadog for monitoring metrics, logs, and alerts through MCP.
ClickUp task management via MCP. Tasks, spaces, lists, goals, and time tracking.
Camunda offers a BPMN-based platform for agent builders to model and automate business processes with decision tables and external task patterns for AI integration. It supports agent-driven process orchestration.
AWS conversational AI service. Tight integration with AWS ecosystem. Good for voice/text bot use cases.
Official MCP server for PostgreSQL. Query, insert, and manage your database from agent workflows. Read-only mode available.
Official Stripe MCP server. Query customers, charges, subscriptions, and products. Full read/write access with careful permission scoping.
Protocol rail for secure agent-mediated commerce flows with explicit checkout and payment confirmation controls.
Google Cloud platform for building and deploying AI agents with support for A2A protocol and agent-to-agent communication patterns.
Elasticsearch search and analytics via MCP. Index documents, run queries, manage indices.
Atlassian Confluence knowledge base via MCP. Search pages, create content, manage spaces.
Enables AI agents to access Cerebras code generation and inference capabilities via MCP.
Cloud browser automation for agents. Managed Chrome instances with session replay, anti-detection, and proxy support.
A fully managed Azure service providing enterprise security, compliance, and Microsoft 365 integration for agent deployment.
Asana task management via MCP. Create tasks, manage projects, update assignees.
Coinbase's framework for onchain agents. Wallet operations, DeFi transactions, blockchain queries.
Enterprise-grade RLHF and annotation platform. 50,000+ vetted reviewers, SOC 2 Type II certified, strong audit trail. Our top pick for enterprise HITL pipelines.
Microsoft's multi-agent conversation framework. Excellent for agent-to-agent delegation patterns. Strong enterprise adoption and active development.
Email marketing via MCP. Manage audiences, create campaigns, query stats.
IBM's enterprise framework for building autonomous agents with native ACP support and semantic multi-agent dialogue capabilities.
High-performance RPC framework using Protocol Buffers for efficient agent-to-agent communication and service invocation.
Google's official agent framework. Native Gemini integration, multi-agent coordination, enterprise tooling.
Official MCP server for Brave Search API. Web and local search with privacy-first results. Reliable, fast, no rate limit surprises on paid plans.
Managed AWS runtime for orchestrating LLM agents, tool calls, memory, and guardrails inside Bedrock workloads.
Behavioral analytics via MCP. Query user journeys, funnels, retention cohorts.
Make.com automation scenarios via MCP. Trigger scenarios, pass data, monitor runs.
Open-source LLM observability platform with tracing, evaluation, and prompt performance analytics.
Offers AI agents access to Jina AI's search, embedding, and multimodal tools via MCP.
CRM operations via MCP. Read contacts, companies, deals, and update pipeline stages. Requires OAuth setup.
MCP server for GitHub repository management and API access. Provides code collaboration tools for development agents.
Programmable AI voice calling platform for outbound and support workflows with optional human escalation paths.
An open protocol enabling agents to securely interact with external tools and data sources through standardized client-server communication.
Provides AI agents with financial market data, stock quotes, and technical indicators via MCP.
Search API built for AI agents. Returns structured, citation-ready results. Faster and more agent-friendly than general search APIs.
Type-safe agent framework from the Pydantic team. Structured outputs, validation, clean Python ergonomics.
Durable execution for agents. Start workflows, signal running executions, query state. Best-in-class reliability for long-running agent tasks.
Durable workflow triggers via MCP. Send events, cancel runs, query function history. Excellent for agent orchestration.
IBM's conversational AI. Mature platform, good for regulated industries. UX is dated.
Secure MCP server for file operations with configurable access controls. Provides agent builders with safe local file read/write capabilities via MCP protocol.
MCP server for fetching web content and converting it for efficient LLM processing. Allows agents to access and parse external web data through standardized MCP tools.
Dagster enables agent builders to build data pipelines as code with asset management and observability for ML and analytics workflows. Agents leverage its API for orchestrated data transformations.
Knowledge graph memory for agents. Stores and retrieves structured agent memories efficiently.
Official Cloudflare MCP server. Manage Workers, KV, R2, D1, and Durable Objects from agent workflows.
CAMEL enables communicative agents for multi-agent role-playing and task-solving through natural language protocols. It supports emergent cooperation in agent workflows.
Autonomous AI agents for enterprise workflows. Pre-built agents for finance, HR, and operations.
The underlying framework powering Microsoft AutoGen agents. Conversation patterns, tool integration.
A research-grade multi-agent framework supporting real-time event-driven orchestration with deep observability into agent decision-making.
Anthropic's tool use specification that enables Claude models to interact with external tools and APIs through structured schemas and JSON-based communication.
Connects AI agents to Supabase for database queries, authentication, and edge functions through MCP.
Official SQLite MCP server. Local database operations with schema introspection. Ideal for local agent state.
Official MCP server for Redis. Get, set, list, and hash operations. Useful for agent state management and caching workflows.
IBM-led protocol for standardized agent communication. REST-based, multi-framework compatible. Companion to MCP for agent runtime communication.
An open-source SDK combining enterprise-ready features with multi-agent orchestration patterns, supporting both LLM-driven and deterministic orchestration with graph-based workflows.
Lightweight remote procedure call protocol specification used by MCP and A2A for standardized message formatting between agents and tools.
Allows AI agents to crawl websites and extract structured data via MCP server.
Google's enterprise conversational AI platform. State-based flows, good NLU. Pricing can be unpredictable.
End-to-end platform with automated pipelines featuring human checkpoints for data labeling. Supports agent builders integrating HITL into full AI operations workflows.
Distributed event streaming platform enabling asynchronous agent-to-agent communication and event-driven agent coordination.
Delivers managed HITL labeling with human review workflows integrated into AWS ML pipelines. Supports agent workflows needing scalable human annotation via AWS APIs.
Agentforce enables Salesforce users to build and orchestrate multi-agent workflows with deep CRM integration and MCP support. Agent builders can deploy voice-enabled agents for sales and service automation.
Enables AI agents to perform file system operations like reading, writing, and listing files through a secure MCP server.
Enables AI agents to control browsers, perform web automation, and testing using Playwright via MCP.
Microsoft's SDK for building AI agents that integrates plugins and tools with structured function calling and semantic understanding.
Official Pinecone vector database MCP. Upsert and query embeddings directly from agent pipelines. Clean schema, reliable uptime.
Make provides a visual automation platform where agent builders connect apps and embed AI modules for intelligent decision-making in workflows. It supports complex scenarios with HTTP modules for agent invocation.
Stateful multi-agent orchestration built on LangChain. The most production-ready open-source A2A framework — graph-based, checkpointing, human-in-loop support.
Provides AI agents access to Firebase services including Firestore, Auth, and Functions via MCP.
Enterprise conversational AI platform. Good for contact center automation, omnichannel deployment.
Enables AI agents to create, edit, and manage Canva designs programmatically via MCP.
Scheduling automation via MCP. Create booking links, check availability, manage event types.
Evaluation and quality platform for testing prompts, models, and agent outputs against production benchmarks.
A framework for building agents with Claude that supports prompt caching for efficient multi-turn conversations and agent memory.
Agno provides primitives for building multi-modal agents with memory and tool integration. It supports scalable agent orchestration for production workflows.
Protocol launched by Google that enables secure cross-organizational agent communication through agent cards, signed messages, and structured metadata for trust and routing.
AWS's open-source agent SDK. Production-tested at Amazon scale. Strong integration with AWS services.
Serverless Postgres via MCP. Branch databases, query with natural language, manage schemas. Good fit for agent-driven development workflows.
AutoGen Studio offers a no-code interface for designing multi-agent topologies with benchmarking. Developers prototype and monitor agent collaborations visually.
LLM observability platform for request tracing, cost tracking, and reliability analytics.
NLP platform with HITL workflows including task management and approval processes. Provides audit trails and versioning for compliance-focused agent applications.
An open-source, low-code platform providing visual building blocks for creating AI agents and multi-agent systems, built on top of LangChain and LlamaIndex.
Role-based multi-agent framework. Define crews of specialized agents with clear responsibilities. Widely adopted, good documentation.
Botpress provides a visual builder for creating conversational agents with multi-channel deployment and knowledge integration. Developers use it for dynamic flows and analytics in agent interactions.
Cloud functions for AI workloads. Run Python agents serverlessly with GPU access. Excellent performance.
Real-time web search with citations via MCP. Research-grade search with source attribution.
Background jobs and scheduled tasks via MCP. Reliable job queuing for agent-triggered async work.
Official browser automation MCP using Puppeteer. Navigate pages, fill forms, take screenshots, extract content.
Official Vercel MCP server. Deploy projects, manage environment variables, inspect deployments and build logs.
Official Slack MCP server. Send messages, read channels, manage threads. Requires workspace OAuth setup.
Prisma ORM operations via MCP. Introspect schemas, run queries, manage migrations from agent workflows.
High-quality data labeling with rigorous QA. Specializes in complex reasoning tasks and safety evaluations. Strong accuracy guarantees.
Ethical research platform with pre-screened participants. Excellent for high-quality HITL data collection where demographic targeting matters.
Personalized memory layer for AI applications. Simple API, cross-session memory, fast retrieval.
Long-term memory management for LLM agents. OS-inspired memory hierarchy. Production service available.
Letta (formerly MemGPT) provides stateful agents with long-term memory management for extended conversations and tasks. It enables persistent multi-turn interactions across sessions.
Prompt injection and unsafe-input protection layer for LLM applications and agent pipelines.
Open-source tool for flexible data labeling with extensible HITL workflows. Agent builders can self-host for custom human review interfaces.
Kubernetes cluster management via MCP. Deployments, pods, services, config maps. High capability, high risk — approval mode required.
Enterprise AI agents for customer and employee experiences. Low-code builder with pre-built industry solutions.
Structured LLM outputs using Pydantic. Simple, reliable, widely-adopted for extraction tasks.
Evaluation and monitoring suite for LLM apps with experiment tracking and prompt performance analysis.
Gumloop provides a no-code interface for agent builders to construct AI workflows using natural language prompts and pre-built nodes for data extraction and automation. It supports agent-like chains for complex task handling.
Query language and schema specification that enables agents to understand and invoke GraphQL APIs with structured type definitions.
Official Anthropic-maintained MCP server for GitHub. Read repos, create issues, manage PRs, search code — full GitHub API surface via clean MCP tools.
Connects AI agents to Jira, Confluence, and Bitbucket for project management via MCP.
Agent-native email identity and inbox infrastructure designed for autonomous workflows and machine-operated communication channels.
Activepieces offers an open-source automation platform for agent builders with a flow-based editor, 100+ integrations, and cloud/self-host options. Agents invoke pieces via API for modular workflows.
Financial data access via MCP. Account balances, transactions, identity verification. High security bar.
Developer-friendly email via MCP. Clean API, React email templates, excellent deliverability. Better DX than legacy email APIs.
Weaviate vector search via MCP. Semantic search, hybrid search, and object management for RAG pipelines.
Creates MCP-compatible context servers from structured/unstructured data sources. Supports RAG pipelines with modular data loaders for agents.
Provides custom workflows with review and consensus stages for human annotation of data in AI training pipelines. Enables agent builders to integrate human quality control into labeling processes via API.
Official Google Drive MCP server. Search, read, and manage files. Requires Google OAuth setup.
Trigger and monitor GitHub Actions workflows via MCP. List runs, check status, download artifacts.
Open-source LLM app development platform. Agent workflows, RAG, prompt management. Fast-growing community.
LLM evaluation framework for deterministic test suites, regressions, and scoring pipelines.
A Python framework for orchestrating role-based autonomous AI agents that collaborate to accomplish complex tasks, featuring sequential and hierarchical process models.
Database platform with automation supporting human review/approval workflows. Agent builders use as backend for HITL task management and routing.
LLM testing and evaluation framework. Red-teaming, regression tests, model comparison. Not a runtime framework.
Smolagents offers minimal code patterns for building agentic workflows with tool calling and state management. It emphasizes simplicity for rapid prototyping of multi-step agent tasks.
Official MongoDB Atlas MCP server. CRUD operations, aggregation pipelines, full-text search from agent workflows.
Qdrant vector database MCP server. High-performance similarity search for agent memory and RAG systems.
Official Notion MCP server maintained by Notion. Search pages, create content, manage databases from agent workflows.
Enterprise platform using weak supervision with HITL for programmatic labeling. Supports agent builders creating scalable labeling functions with human validation.
Knowledge graph-based persistent memory system via MCP. Allows agents to store, retrieve, and query long-term memory across sessions.
Magentic-One is a production-ready multi-agent framework with Docker deployment and observability. It orchestrates specialized agents for collaborative task execution.
A data framework for building RAG applications and agent systems with support for data indexing, retrieval, and integration with multiple LLM providers.
A stateful orchestration framework within the LangChain ecosystem for building complex agent workflows with graph-based control flow, persistent state, and checkpointing for long-running tasks.
A visual, graph-based framework for orchestrating LLM workflows with a cyclical architecture supporting shared memory and multi-agent systems.
Kestra is an open-source orchestration platform for agent builders to define workflows as code with YAML, supporting AI tasks and 100+ plugins. It enables scalable, agent-triggered data and ML pipelines.
Atlassian Jira via MCP. Create issues, manage sprints, query projects. Strong feature coverage.
Allows AI agents to deploy and manage static sites on EdgeOne Pages via MCP protocol.
Open platform co-developed with IBM for managing the agent lifecycle including installation, running, registration, and sharing of agents with offline agent discovery.
Anthropic's official tool use, computer use, and agent patterns. Reference implementation for Claude-based agents.
The original crowdsourcing marketplace. Massive scale but requires significant QA overhead to achieve acceptable quality.
REST-based protocol developed by IBM Research that enables autonomous AI agents to communicate and coordinate with one another using structured messages and semantic dialogue.
Distributed SQLite via MCP. Create databases, run queries, manage schemas at the edge. Good for multi-tenant agent workflows.
Product analytics via MCP. Query events, feature flags, session recordings, experiments.
A flexible, code-first framework for building AI applications with support for multiple languages (Python, JavaScript) and custom agent architectures.
A pipeline-first framework for building RAG systems and agent workflows with DAG pipelines, function-calling tools, and hybrid search capabilities.
Role-based agent framework. Clear task delegation model. Better documentation than most open-source alternatives.
Software engineering agent focused on end-to-end coding tasks across repositories, tooling, and execution loops.
LLM tracing and evaluation suite for observability, root-cause debugging, and model quality monitoring.
Advanced Message Queuing Protocol for reliable agent-to-agent messaging and task coordination in distributed systems.
Open-source automation platform via MCP. Trigger flows, pass structured data, get execution results.
Structured text generation library. JSON/regex-constrained outputs. Fast and reliable structured generation.
Long-term memory store for LLM apps. Temporal knowledge graphs, fact extraction, session management.
Rasa enables development of conversational agents with multi-agent orchestration for collaborative tasks and dialogue management. It integrates NLU, entity extraction, and channel support for context-aware interactions.
Phidata (Pydantic AI Data) enables building agents with tools, memory, and vector stores using structured data models. It supports multi-agent workflows through composable agent templates.
Run ML models via API. Not an agent framework per se, but widely used for model-calling in agent pipelines.
Official Shopify MCP server. Products, orders, customers, inventory management from agent workflows.
SMS, voice, and WhatsApp via MCP. Full Twilio API surface for agent-driven communications. Approval mode recommended for outbound messages.
Query Sentry errors, performance data, and releases. Useful for agents involved in incident response.
Serverless MySQL via MCP. Branch databases, deploy schema changes, query with connection pooling.
Computer vision annotation with auto-labeling. Good for image and video datasets with complex annotation requirements.
Open annotation platform for computer vision. Strong community, good preprocessing tools. Free tier is generous.
Dataset management and curation platform. Find edge cases, track model performance, manage annotation queues.
Supports scriptable annotation workflows with active learning loops for NLP tasks. Allows agent builders to create local HITL feedback mechanisms for model improvement.
Extract, transform, and load structured data across databases and APIs. Handles schema validation and error recovery.
Emerging standard for structured handoffs from autonomous agents to human reviewers. Covers escalation triggers, context packaging, and response routing.
Standard for describing asynchronous APIs and event-driven architectures, enabling agents to understand and invoke async services.
AI integration framework for Java/Spring. Enterprise Java ecosystem integration for LLM applications.
Send transactional emails via MCP. Template management, contact list operations, delivery tracking.
Open-source workflow platform via MCP. Run scripts, trigger flows, manage resources.
Customer data platform via MCP. Track events, query profiles, manage audiences.
AI engineering toolkit by Prefect. Functions, classifiers, extractors built on LLMs. Pythonic API.
Automation platform supporting complex HITL workflows with human approval steps. No-code interface for agent orchestration with review gates.
Active learning platform for computer vision with collaborative HITL annotation. Enables agent builders to iterate models with human feedback loops.
Open-source computer vision annotation tool with team review workflows. Self-hosted option for agent builders needing custom HITL interfaces.
Open-source embedding database for agent memory. Simple API, local-first. Good for prototyping RAG systems.
Allows AI agents to execute SQL queries and manage SQLite databases directly via MCP protocol.
Microsoft's TypeScript-first structured LLM output library. Type-safe schema-validated responses.
Fast inference for open-source models. Good for agent pipelines that need cheap, fast LLM calls.
Low-code agent builder integrated with Microsoft 365. Good for enterprise M365 environments.
No-code AI workforce builder. Create agents with tools, memory, and reasoning. Good for non-technical teams.
Customer messaging platform via MCP. Conversations, contacts, articles, events.
Salesforce CRM via MCP. SOQL queries, object CRUD, workflow triggers. Enterprise-grade but complex setup.
Incident management via MCP. Trigger alerts, acknowledge incidents, manage escalations.
Customer support ticket management via MCP. Create tickets, update status, search conversations.
Framework for integrating tools and APIs with language models, supporting structured tool definitions and agent-based tool invocation.
Provides AI agents with tools to access GitHub repositories, retrieve commits, issues, and perform repository operations via MCP protocol.
Schedule and manage Zoom meetings via MCP. Create meetings, retrieve recordings, manage webinars.
Personalize onboarding sequences, provision resources, schedule check-ins, track milestone completion.
Multi-agent workflow: Tier-1 AI handles common queries, auto-escalates to human queue when confidence drops below threshold.
Transcribe meetings, extract decisions and action items, create follow-up tasks in project management tools.
Offers mathematical computation tools for AI agents to perform calculations through MCP server.
Official reference implementation for local filesystem access. Read, write, list, and search files. The canonical MCP server for file operations.
Modern labeling platform emphasizing quality control workflows and reviewer consensus. Agent builders can use APIs for human verification of AI outputs.
Open-source validation and policy framework for enforcing output quality and safety constraints in LLM systems.
HITL solution for human oversight in AI workflows with webhook callbacks for responses. Framework-agnostic SDKs enable agents to pause for team review.
EmbedChain simplifies building RAG agents by handling data ingestion, indexing, and querying in multi-step pipelines. It supports agent orchestration with custom tools and memory persistence.
SuperAGI provides a platform for building and deploying autonomous agents with toolkits and action consoles for workflow management. It supports multi-agent collaboration through hierarchical structures.
Open-source visual AI programming environment by Ironclad. Good for complex prompt chains.
Conversation design platform for AI agents. Good for voice and chat experiences. Less suitable for complex backend automation.
Webflow CMS and site management via MCP. Query collections, create items, manage pages.
Trigger n8n workflows via MCP. Execute webhooks, pass data to automation pipelines, retrieve results.
Form and survey management via MCP. Create forms, retrieve responses, manage response data.
Accounting and invoicing via MCP. Create invoices, query accounts, manage contacts.
Open protocol/network initiative focused on interoperable multi-agent communication and discovery.
Official knowledge graph memory server. Store and retrieve agent memories across sessions.
Enterprise AI training data with multilingual capabilities. Strong for localization-sensitive tasks.
Langroid supports multi-agent conversations in a single process with task delegation. It enables developers to create collaborative agent teams for complex reasoning.
Provides AI agents with recipe data and cooking instructions through MCP protocol.
MCP server offering tools to read, search, and manipulate Git repositories. Enables code agents to interact with version control systems programmatically.
Official HTTP fetch MCP server. Make GET/POST requests, convert HTML to markdown. Core utility for web-connected agents.
Managed workforce for AI data labeling. Good project management overhead, consistent quality on structured tasks.
ChatDev simulates a software company with multi-agent collaboration for full development cycles. Agents communicate as CEO, CTO, and programmers to build applications.
Research framework for communicative multi-agent systems. Role-playing, task-oriented conversations between agents.
Crowdsourced data annotation platform with HITL quality control for AI training data. Agent builders access managed human labeling through APIs.
Microsoft's tool for LLM application development and evaluation. Visual flows, CI/CD integration.
A lightweight, experimental open-source multi-agent orchestration framework from OpenAI for building simple agent coordination patterns.
PayPal payment operations via MCP. Create orders, capture payments, manage subscriptions.
AI training data and content moderation. Enterprise contracts, strong compliance posture.
Human insight platform for UX and AI output evaluation. Good for qualitative HITL tasks.
AI agent for customer support built on Intercom. Excellent for support use cases, limited to Intercom ecosystem.
Framework for building agents with tool integration capabilities using Hugging Face models and structured tool definitions.
Open-source testing framework for LLM and ML systems including safety scans and quality checks.
Trello board management via MCP. Cards, lists, boards, attachments.
Intuit QuickBooks accounting via MCP. Invoices, expenses, customers, reports.
Miro whiteboard via MCP. Create sticky notes, shapes, connectors, and read board content.
Gather competitor pricing, product features, and positioning from public sources. Synthesize into structured report.
Programming language for LLMs. Constraint-based decoding, Python integration. Research-oriented.
Customer support AI agents for enterprise service operations with workflow and system integrations.
Automated agents built on Zapier's 7000+ app integrations. Good for non-technical users.
Native AI agents within Zendesk. Triage tickets, resolve Tier-1 issues. Locked to Zendesk ecosystem.
Monday.com work OS via MCP. Boards, items, columns, automations.
Active learning and data curation for computer vision. Reduces labeling cost via smart sample selection.
Evaluation platform combining LLM judges with human-in-the-loop review workflows. Agent builders configure custom forms and reviewer routing for output validation.
AI employees for go-to-market operations with autonomous outbound execution and workflow automation.
Open-source data annotation framework for NLP and LLMs. Self-hostable, developer-friendly, strong community.
Observability and monitoring for AI agents. Session tracking, cost analysis, error debugging.
Text and document annotation platform with OCR and NLP capabilities. Good for invoice and contract processing.
Monitor brand mentions, classify sentiment, alert on critical issues, draft response suggestions.
API and SDK for integrating human decision-making into AI agent workflows with multi-channel routing. Allows agents to request human approval via Slack, email, SMS, or WhatsApp.
WordPress REST API via MCP. Posts, pages, media, users, taxonomies.
Tracing, testing, and evaluation platform for LLM applications and agent workflows built by LangChain.
Read and send Gmail via MCP. Thread management, label operations, search. Sensitive — approval mode strongly recommended.
Crowd tasking platform for human verification and data labeling tasks. Enables agent builders to distribute HITL tasks to global workforce via API.
Bardeen provides AI-powered workflow automation for agent builders with browser extensions and playbooks that integrate apps via natural language commands. It enables autonomous agent actions across web apps.
AirOps equips agent builders with Power Agents for SEO and content workflows, integrating tools like Semrush for automated research and generation tasks. It supports scalable agent-driven content pipelines.
SuperAGI is a dev-first platform for autonomous agents with multi-agent collaboration features. It provides tools for action kits and performance telemetry in agent fleets.
Platform for human annotation tasks including computer vision and NLP labeling. Supports agent workflows requiring on-demand human review.
Mage AI is an open-source platform for agent builders to develop, test, and deploy data pipelines with Python blocks and LLM integrations. It supports agent scheduling and orchestration via API.
Video messaging via MCP. Search recordings, get transcripts, create sharing links.
Personal AI agent for email, calendar, and meeting workflows. Non-technical user focus. Limited API access.
Crowdsourcing platform for data labeling and human intelligence tasks. Provides APIs for agent builders to integrate HITL quality assurance.
AI-powered customer service automation. Strong in Asian markets. Multilingual NLU.
Developer-first AI assistant and agent builder. Clean API, good custom tool integration. Small but active team.
AI agent hosting platform. Package and deploy agents as persistent services.
Minimal task-driven autonomous agent. Research/educational value. Not suitable for production.
DSPy is a programming model for optimizing LM prompts and weights in agentic pipelines. It structures multi-step reasoning workflows through declarative modules.
AgentVerse is a platform for creating and simulating multi-agent environments with shared communication protocols. It facilitates complex interactions and emergent behaviors in agent ecosystems.
Multi-agent project management and automation. Promising workflows but scoring data has aged — flagged for re-review.
Standard for validating and describing JSON data structures used by agents to understand tool inputs, outputs, and API schemas.
RevOps automation agents for outbound, pipeline workflows, and sales execution operations.
MetaGPT implements a multi-agent framework simulating a software company with roles like PM and engineer for collaborative coding workflows. It generates code and documentation through structured agent interactions.
Communicative agents for role-playing and task solving. Research-origin framework with growing production adoption.
Lindy allows agent builders to create AI agents for workflow automation with natural language setup and multi-agent coordination. It integrates common apps for autonomous task execution.
Huginn is an open-source system for agent builders to create agents that monitor events and perform automated tasks across websites and services. It supports custom agent development for workflow extension.
AI-powered customer service automation. Pre-built integrations with helpdesk platforms. Good accuracy on FAQ tasks.
Domain-focused legal AI agent platform for drafting, research, and legal workflow acceleration.
Reference MCP server providing prompts, resources, and tools for testing protocol capabilities. Enables agent builders to experiment with core MCP features like tool invocation and resource access.
Medical AI annotation with clinical expert reviewers. Specialized for healthcare — not suitable for general tasks.
Classify incoming emails, extract action items, route to appropriate team queues, draft responses.
AI-powered data labeling with human QA. Fast throughput, competitive pricing for image and video tasks.
AutoGPT facilitates autonomous agents with multi-agent extensions for task decomposition and execution. Builders configure plugins for agent-to-agent coordination.
Microsoft's constraint-based generation library. Control LLM output structure precisely.
Developer-first annotation platform. Good API-first design for integrating HITL into ML pipelines.
Model evaluation and HITL review platform built for AI teams. Clean API for routing tasks to human evaluators, with native CI/CD integration.
Developer-focused RLHF annotation service. Annotators with coding backgrounds for technical task evaluation.
Automatisch is an open-source Zapier alternative for agent builders to self-host workflow automations with 100+ app integrations. It supports custom triggers and actions for agent-compatible pipelines.
Developer platform for AI agents with tool use. API-first design. Small team, uncertain roadmap.
AI-assisted annotation with managed human workforce. Multi-modal support. Documentation is limited.
Team-based text annotation platform. Simple UX, good inter-annotator agreement tracking.
Browse AI helps agent builders extract structured data from websites via no-code robots that can be integrated into workflows as API endpoints. Agents use it for web scraping and monitoring tasks.
Autonomous Agents framework supports multi-agent collaboration with customizable roles and tool integration. It facilitates agent communication for research and task automation.
Data annotation platform with HITL workflows for training AI models. Offers APIs for scalable human review in agent data pipelines.
Camphr extends spaCy with agent capabilities for NLP pipelines and multi-step processing workflows. It supports custom components for agentic text analysis tasks.
Analyze experiment results, calculate statistical significance, generate recommendations.
Analyze codebases, generate API documentation, create usage examples, maintain changelogs.
Extract data from invoices, validate against PO, route for approval, trigger payment. Requires HITL approval gate.
Review PRs for security issues, style violations, and logic errors. Integrates with GitHub MCP.
Browser-based autonomous agent interface. Easy to try, not production-grade. Useful for demos.
Analyze incoming bug reports, reproduce with code interpreter, assign severity, create Jira tickets.
Research prospects, score against ICP criteria, draft personalized outreach. Integrates with CRM MCP servers.
Scrape competitor sites for pricing, features, and announcements. Weekly digest with structured diff.
Research a topic across multiple sources, synthesize findings, draft outline. Uses search + fetch MCP servers.
The HTTP API specification standard. Version 3.1 adds JSON Schema 2020-12 alignment. Used as the base description layer for most MCP server definitions.
Anthropic-originated open standard for connecting AI models to tools and data sources. Now maintained by the Linux Foundation's Agentic AI Foundation. 75+ connectors across major AI platforms.
Industry standard for describing REST APIs that enables AI agents to understand and invoke API endpoints through structured schema definitions.
Standard for composable WebAssembly components enabling agents to load and invoke tools as portable, sandboxed modules.
Bidirectional communication protocol enabling real-time agent-to-agent messaging and persistent connections for agent coordination.
Google's method for serializing structured data used in gRPC and other agent communication protocols for efficient message encoding.
Node-RED provides a flow-based programming tool for agent builders to wire together hardware, APIs, and services with JavaScript nodes including LLM support. It runs agent workflows locally or on edge devices.
Interactive documentation and testing tool for REST APIs that helps agents understand and discover available tools and endpoints.
Temporal provides a durable execution platform for agent builders to write reliable, long-running workflows in code with state management and retries. Agents use it for complex, stateful automations.
Prefect is a workflow orchestration tool for agent builders to schedule, monitor, and retry Python workflows with hybrid execution modes. It integrates LLMs for dynamic agent workflows in data engineering.
Open standard from Anthropic that standardizes how AI agents connect to business tools and data sources through a universal interface using JSON-RPC 2.0 communication.
OpenAI's specification for enabling language models to call external functions and APIs with structured JSON schemas and natural language descriptions.
Python library for defining and validating tool schemas using type hints, enabling structured tool definitions for AI agents.
MCP server providing time and timezone conversion capabilities. Enables agents to handle temporal data and scheduling across different regions.
MCP server for querying and analyzing MongoDB collections. Provides agent builders with database access tools for NoSQL data operations.
MCP server integrating MySQL databases with configurable access controls and schema inspection. Enables SQL database interactions for data-driven agents.
MCP server for natural language task management in Todoist. Enables agents to create, update, and query tasks programmatically.
Exposes Notion pages, databases, and tasks as MCP context for LLMs. Allows knowledge agents to reference workspace data in real-time.
Bridges Supabase Postgres and edge functions to stream contextual data to LLMs. Provides serverless scalable context delivery for agents.
MCP server providing access to Redis key-value stores through standardized tools. Enables fast caching and session storage for agents.
MCP server connector for Perplexity API enabling web search. Allows agents to perform research without leaving MCP ecosystem.
MCP server providing browser automation capabilities via Playwright. Enables agents to interact with web applications programmatically.
MCP server integrating Sentry.io for error tracking and performance monitoring. Enables agents to monitor application health and debug issues.
MCP server for interacting with Postman API collections and environments. Allows agents to manage and execute API tests programmatically.
Captures real-time Slack conversation threads and metadata for LLMs. Enables enterprise bots to provide context-aware responses in channels.
MCP server connecting AI tools to 500+ apps including Gmail, Slack, GitHub, Notion. Provides unified access to multiple SaaS platforms for agents.
Enables AI agents to perform web searches and retrieve search results through MCP protocol.
Connects AI agents to MiniMax AI services for generation and processing tasks via MCP.
Allows AI agents to scaffold, build, run, and test iOS/macOS apps with UI automation via MCP.
Provides AI agents access to monitoring data, uptime status, and incident management via MCP.
Provides AI agents with web search capabilities through Zhipu search service via MCP.
Provides AI agents with advisory tools and recommendations through MCP server.
Offers managed human labeling services with precision review workflows for high-accuracy datasets. Agent builders can submit tasks for human verification through APIs.
Provides computer vision annotation tools with productivity-focused HITL workflows. Enables teams to route agent-generated labels for human correction via web and API.
Vision-first platform with automated labeling and human review workflows. Provides APIs for agent builders to incorporate HITL in computer vision pipelines.
MCP server for dynamic reflective problem-solving through structured thought sequences. Helps agents break down complex tasks into manageable reasoning steps.
Zapier allows agent builders to automate multi-step workflows with built-in AI agents, MCP support, and over 8000 app integrations via drag-and-drop or code steps. Agents can be triggered programmatically for autonomous operation.
n8n enables agent builders to create workflows using a visual node-based editor with native LLM and AI agent nodes for tasks like data processing and API orchestration. It supports self-hosting and cloud deployment for flexible agent integration.
Vellum provides an all-in-one platform for agent builders to experiment, deploy, and monitor complex AI agents with workflow orchestration capabilities. It supports enterprise-scale automation with analysis tools.
Vertex AI Agent Builder allows creating enterprise agents with multi-agent orchestration via Google's cloud platform. Developers integrate grounding data and tools for production agents.
TaskWeaver is a code-first agent framework generating executable code for task orchestration. It enables multi-agent collaboration through planning and verification modules.
A multi-language orchestration platform supporting .NET, Python, and Java for building production-grade agent systems with skill-based planning and Azure AI integration.
A platform for developing, deploying, and optimizing agent workflows combining visual Agent Builder with managed deployment options and code export capabilities.
An enterprise platform combining AI agents, Maestro orchestration with BPMN-based workflow modeling, and process intelligence for coordinating agents, RPA bots, and human reviewers.
Lightweight publish-subscribe messaging protocol used for agent-to-agent communication in IoT and distributed agent systems.
A combination of Prefect workflow orchestration with ControlFlow extension for orchestrating LLM agents with production-grade features like scheduling, retries, and observability.
An MLOps framework for orchestrating LLM workflows and agent pipelines with production-grade features for deployment and monitoring.
An enterprise-focused platform for building AI agents with compliance and security features for regulated industries.
A visual, low-code platform for building agent workflows with integration to thousands of applications and APIs.
Pipedream lets developers build AI agents using natural language with thousands of API integrations and serverless workflows. Agent builders can create custom MCP servers for extensible connectivity.
Windmill offers an open-source platform for agent builders to script workflows in TypeScript, Python, or Go with native AI job support and low-code components. It enables self-hosted agent orchestration with granular permissions.
Workato delivers enterprise-grade automation for agent builders with MCP, prebuilt AI agents (Genies), and 1200+ integrations plus governance features. It supports multi-step agent orchestration with RBAC.
Stack AI offers a low-code platform for agent builders to design multi-step AI workflows with drag-and-drop nodes for LLM chaining and app integrations. It facilitates rapid agent prototyping and deployment.
Relay.app enables agent builders to automate workflows with human-in-the-loop AI agents using natural language instructions and 100+ integrations. It supports programmatic triggering for agent sequences.
Stripe's official MCP-oriented toolkit for agent-triggered payments and billing operations with approval controls.
Prompt management and observability platform for versioning, tracing, and evaluating production prompt changes.
LLM application tracing and evaluation toolkit integrated with experiment tracking workflows.
AI gateway and observability layer with routing, guardrails, and evaluation tooling for production agents.
Prompt injection detection toolkit for agent systems with defensive checks for untrusted content.
Safety and evaluation platform focused on LLM quality assurance, hallucination detection, and production guardrails.
NVIDIA guardrails framework for controlling conversational behavior, policy compliance, and safe tool use.
Salesforce enterprise agent platform for building and operating autonomous and assistive agents across CRM workflows.
AI-native documentation infrastructure with machine-readable docs surfaces and MCP-oriented integration patterns.
Open-source telemetry and observability stack for LLM applications and agent pipelines.
Financial rails for autonomous agents, focused on compliant machine-to-machine payments and identity-aware transactions.
Agent payment infrastructure for authorizing, routing, and auditing automated transaction flows.
Open-source evaluation framework for RAG and agent outputs with quality metrics and benchmark workflows.
Enterprise customer experience agent platform for high-volume support and commerce conversations.
Voice agent platform for phone-based support and operations with enterprise controls and monitoring.