REPO PACKAGE
langchain-ai deepagentsjs
COMMUNITY7skills
3ready to use
LABSCO SUMMARY
Only eval-writer is scoped to this codebase: it scaffolds new eval suites for the deepagentsjs monorepo itself — dataset design, vitest configuration, LangSmith integration — for someone contributing test coverage to the project. langgraph-docs fetches current LangGraph documentation on demand, and langsmith-trace adds tracing to an application or queries existing traces through the langsmith CLI, both aimed at someone building with LangChain's own tools rather than with deepagentsjs specifically. The remaining four — arxiv-search, file-organizer, skill-creator, and web-research — are generic capabilities with no reference to deep agents, LangChain, or this repository anywhere in their own descriptions: searching arXiv, tidying local files, writing a new skill, and running a structured web-research pass.
Someone installing deepagentsjs for its actual purpose — createDeepAgent, the planning, filesystem, and sub-agent harness the README describes — gets nothing from any of the seven; they are unrelated tooling that happens to live in the same repository, not skills for using the package. eval-writer is useful only to a contributor working on deepagentsjs's own test suite, and the other six are general-purpose skills any agent could use regardless of whether deepagents is involved at all.
READ THE FULL ANALYSIS
What each needs to run. Three of the seven — file-organizer, langgraph-docs, and web-research — work with nothing extra installed. skill-creator and arxiv-search need a local tool already on the machine. eval-writer and langsmith-trace both assume a stored LangSmith account key, since tracing and eval scoring both run through LangSmith.
A separate Python package exists too. The README points explicitly to a sibling repository, langchain-ai/deepagents, for anyone wanting the same harness outside JavaScript or TypeScript — a different repository, not covered by this page.
WHAT'S INSIDE
7 showing · 7 totalReady to use
Nothing else to set up — install it and go.
7 of 7
SETUP
Showing all 7 skills
arxiv-search
arXiv is a free public archive where researchers post papers in science, maths and computing before any journal prints them; this searches it by keyword and hands back the matching titles and abstracts.
eval-writer
Builds a scored test suite for an AI agent in the deepagents JavaScript project, so you can tell whether a change made the agent better or worse.
file-organizer
A tidy-up helper for a cluttered computer: it reads through a folder, works out what belongs together, points out the duplicate copies, and moves nothing until you have approved the plan.
langgraph-docs
Stops the assistant answering LangGraph questions from memory: it reads the current official LangGraph documentation first, then answers from what it found there.
langsmith-trace
Every call your app makes to a language model gets written down in LangSmith, and those records can then be searched, filtered and exported when something comes out wrong or runs slowly.
skill-creator
How to build your own skill — a folder of instructions that teaches an assistant one particular job — starting from a template and checking it for mistakes before you install it.
web-research
Break a big question into a few smaller ones, send a separate helper off to search the web for each, and pull what they find back into one written answer with its sources.
No skill in this package matches that.
HOW TO GET IT
WHOLE PACKAGEInstall the whole package (7 skills)
npx skills add langchain-ai/deepagentsjsSINGLE SKILLOr install a single skill
npx skills add langchain-ai/deepagentsjs --skill <name> --full-depthPick the skill name from the Skills tab — each entry there installs independently.