@halo-format/langgraph

Halo host adapter for LangChain agents and LangGraph. installHalo() wires it in one call:

• a wrapToolCall encode middleware — the deterministic wrap-the-tool-call hook, LangChain's analog of the Claude SDK's PostToolUse — that replaces a large tool result with a halo shape map (root kind + one line per field: ref, kind, and a bounded preview), so the payload stays out of the model's context while it still sees what's there. The full envelope rides in the ToolMessage artifact (kept in graph state, never sent to the model) for audit/replay;
• a single plain LangChain halo_fetch tool the model uses to pull back only the leaves it needs, verified on read — a ref that lands on a branch returns that branch's sub-refs, so one batch API both pulls and expands (there is no separate halo_walk).

import { createAgent } from "langchain";
import { installHalo } from "@halo-format/langgraph";

const { tools, middleware, session } = installHalo({ tools: myTools });
const agent = createAgent({ model, tools, middleware });
// session holds the shared store for audit/inspection

The middleware is deterministic plumbing (it always fires, for every tool); the Halo Skill (or prompt-mode guidance) is the navigation behavior. Pass store: new FileStore(dir) for the heavy/persistent deployment. The core engine is @halo-format/halo; this package is only the shim.

JS vs. Python note. The JS ToolNode does not accept a wrap option, so on this host the middleware is the only clean interception surface — there is no haloToolNode() the way there is in the Python adapter.