Ross ROSS = Recommend OSS · open-source software intelligence for agents

langchain-ai/deepagents

The batteries-included agent harness. observed · 2026-08-28

github.com/langchain-ai/deepagents · homepage · Python · MIT (permissive) observed · 2026-08-28

Health v2 · maintenance only

81/100

  • Activity 99
  • Release rhythm 87
  • Longevity 28
How is this computed?

round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10) — computed 2026-09-03. Adoption (stars, forks) is never an input.

  • gap_med: 0
  • age_days: 402
  • days_rel: 7
  • days_push: 7
  • n_releases_24m: 246

Full methodology

Adoption not part of the score

28559 stars · 4000 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

Deep Agents is an open-source, batteries-included agent harness from LangChain built on LangGraph, providing an opinionated agent with filesystem tools, sub-agent delegation, context management, persistent memory, and human-in-the-loop approval out of the box. It is model-agnostic (works with any tool-calling LLM), available in Python and TypeScript, and every component can be extended, overridden, or replaced.

Use cases

  • build an AI agent that can plan and execute multi-step tasks
  • delegate subtasks to subagents with isolated context windows
  • give an LLM agent file read/write/search and shell access
  • build a research agent that writes reports
  • add human approval before an agent runs tools
  • run a coding agent in the terminal powered by any LLM
  • persist agent memory across sessions
  • connect an agent to MCP servers and custom tools

When to choose

  • you want a production-ready agent with sensible defaults for long-horizon, multi-step work
  • you need subagents, filesystem tools, context summarization, and memory without building them yourself
  • you want model-agnostic agents that work with OpenAI, Anthropic, Google, or local models
  • you're already in the LangChain/LangGraph ecosystem and want tracing and deployment via LangSmith

When to avoid

  • you need a minimal, unopinionated agent loop and prefer to assemble everything yourself
  • your LLM does not support tool calling
  • you want a fully managed no-code agent builder rather than a code framework
  • you need a framework outside the LangChain ecosystem with no LangGraph dependency

Facets

framework · maturity active

agent-framework llm-inference mcp file-system chatbot large-language-models developer-tools python cross-platform cli agent-harness subagents context-management langgraph human-in-the-loop long-horizon-tasks skills model-agnostic ai-agents automation nodejs

10 sources

Member repositories

RepositoryRoleHealth v2
langchain-ai/deepagentsmain81

For agents

markdown · JSON · MCP: product_card(name="langchain-ai/deepagents")

Data as of 2026-08-30T08:39:29.467469+00:00 · Report a problem