Research: deep-agents-core
Cached research evidence for deep-agents-core (not authority).
Purpose
Section titled “Purpose”Part of official langchain-ai/langchain-skills (early dev). Agent architecture, harness setup, and SKILL.md format for Deep Agents. See README for LangChain/LangGraph/Deep Agents selection via primer.
Harness Coverage
Section titled “Harness Coverage”Target agents: antigravity etc (curated group install). Requires API keys.
Trust And Risks
Section titled “Trust And Risks”trust_tier=needs-inspection; status=inspect-then-install; policy=inspect source/hooks/credentials before install. Official org, early dev (APIs may change), broad scope for agent building implies generated code + tool use risks.
Install Prerequisites
Section titled “Install Prerequisites”Group install via langchain-ai/langchain-skills npx … —skill deep-agents-* etc; status=inspect-then-install.
Upstream Maintainer
Section titled “Upstream Maintainer”Comparable Alternatives
Section titled “Comparable Alternatives”Custom deep agent skills or langgraph equivalents.
> Evidence from public web sources, GitHub (langchain-ai org Jun 2026), curated; research evidence only, not authority/endorsement.