Agent runtimes
Hermes Agent
Provider-agnostic agent with tools, memory, skills, cron, and messaging connectors.
Reference
Tools
Curated tools for building, evaluating, and operating practical AI agent systems.
⌕
Coding agents
Claude Code
Terminal coding agent for repo-level software tasks.
Coding agents
Codex CLI
OpenAI coding agent for terminal and Git workflows.
Protocols
MCP SDK
Reference SDKs for building Model Context Protocol servers and clients.
Orchestration
LangGraph
Graph-based orchestration for stateful agent workflows.
Observability
OpenTelemetry
Trace substrate for agent runs, tool calls, and production observability.
Observability
Langfuse
Open-source tracing and evals for LLM applications.
Evaluation
Promptfoo
CLI eval runner that fits into CI for prompt and agent regression tests.
Schemas
Zod
Runtime schema validation for TypeScript tool definitions.
Schemas
Pydantic
Typed schemas for Python tool inputs and outputs.
Sandboxes
E2B
Sandboxed runtime for executing agent-generated code.
Sandboxes
Modal
Serverless containers for long agent jobs and GPU tasks.