Research: langchain-middleware
Cached research evidence for langchain-middleware (not authority).
Purpose
Section titled “Purpose”Part of official langchain-ai/langchain-skills (early dev, ~800 stars). ecosystem-primer: start-here for choosing LangChain vs LangGraph vs Deep Agents, env setup, next skills. langchain-fundamentals: agents via create_agent, tools, structured output, middleware. langchain-middleware: HITL approval, custom middleware, Command resume. langchain-rag: RAG (loaders, embeddings, vector stores). langchain-dependencies: version pinning ref for py/ts.
Harness Coverage
Section titled “Harness Coverage”Targets from curated: antigravity, claude-code, codex, crush, cursor, gemini-cli, github-copilot, grok, opencode. Requires OPENAI/ANTHROPIC keys for use per README.
Trust And Risks
Section titled “Trust And Risks”trust_tier=needs-inspection; status=inspect-then-install; provenance=verified npx install cmd in config; policy=Inspect source, hooks, scripts, credentials, dedupe before install (per catalog). Early dev warning in repo; LangChain org official but skills in flux. Risks: broad API key needs, potential dep drift, agent-generated code exec patterns in fundamentals.
Install Prerequisites
Section titled “Install Prerequisites”Install (group): npx skills add langchain-ai/langchain-skills —skill ecosystem-primer —skill langchain-fundamentals —skill langgraph-fundamentals … —skill langchain-rag —skill deep-agents-* … -y -g -a [agents]; status=inspect-then-install; selector=named. Set API keys post-install.
Upstream Maintainer
Section titled “Upstream Maintainer”langchain-ai/langchain-skills (official LangChain AI org).
Comparable Alternatives
Section titled “Comparable Alternatives”langchain docs or custom agent building skills; other RAG/LLM framework skills.
> Evidence from public web sources, GitHub repositories (official orgs Jun 2026), curated catalog; treat as research evidence, not authority or endorsement. Use /review source + live audit for promotion.