Build teams of AI agents that collaborate, automate workflows, and complete complex tasks together.
HasteKit
FreeStarting price
0Popularity
About HasteKit
HasteKit provides a production-grade stack for developing and operating LLM-based agents. It includes an LLM gateway that routes requests across multiple providers without client-side changes, durable runtimes for long-running agent loops, and a configuration interface for tuning model behavior, memory, and output formats. The platform supports multi-agent orchestration, allowing teams of specialized agents to collaborate via handoffs or tool calls. It offers reusable skills, knowledge bases with first-class RAG, and built-in tools such as image generation and code execution. Connectors integrate with external services like Slack and GitHub, while workflows enable deterministic DAG-based automation. Channels support user interactions in Slack and Telegram, and triggers automate scheduled or event-driven agent activation. End-to-end observability is built in, with OpenTelemetry tracing and cost tracking per request.
Key features
- LLM gateway with multi-provider routing
- Durable execution on Temporal or Restate
- Multi-agent orchestration with handoffs
- Reusable skills packaged as SKILL.md bundles
- Knowledge bases with first-class RAG
- Built-in tools for image generation, speech, and code execution
- Connectors for Gmail, Slack, GitHub, and others
- Deterministic workflows with 30+ node types
Use cases
- Building and deploying production-grade LLM agents
- Automating multi-step workflows with deterministic execution
- Integrating agents into Slack and Telegram for user interaction
Pros
- Single SDK and endpoint for LLM gateway, agents, and workflows
- Durable execution with automatic retries and crash recovery
- Multi-provider LLM gateway with OpenAI compatibility
- Built-in tools and connectors for common use cases
- End-to-end observability with OpenTelemetry tracing
Cons
- No explicit mention of a free tier or pricing tiers
- Limited to providers and connectors listed on the site
- Requires integration with external services for some features
Frequently asked questions about HasteKit
What is HasteKit?
HasteKit is a production-grade platform for developing and operating LLM-based agents. It provides an LLM gateway, durable runtimes, multi-agent orchestration, skills, RAG-based knowledge bases, built-in tools, connectors, workflows, channels, triggers, and end-to-end observability.
Who should use HasteKit?
HasteKit is designed for teams building and scaling agentic applications, including developers, product teams, and operations staff who need a unified platform for agent development, deployment, and monitoring.
How does the LLM gateway work?
The LLM gateway acts as a drop-in replacement for the OpenAI API, routing requests across multiple providers using a single virtual key. It supports OpenTelemetry tracing, cost tracking per request, and per-project rate limits without requiring client-side changes.
Can HasteKit integrate with external services?
Yes, HasteKit includes first-class connectors for services like Slack, GitHub, Gmail, and Jira. These connectors wrap curated actions as tools and use per-user OAuth, ensuring agents never handle user tokens directly.
Does HasteKit support multi-agent orchestration?
Yes, HasteKit allows agents to collaborate through sub-agent calls, isolated or shared context, or full conversation hand-offs. Teams can compose focused agent groups instead of relying on monolithic prompts.
What observability features does HasteKit provide?
HasteKit includes end-to-end observability with OpenTelemetry tracing and cost tracking per request. Every step in agent execution, including tool calls and memory updates, is logged with full context and citations.
