**TL;DR:** A subgraph compiled with checkpointer=True (continuations mode) cannot be called multiple times inside one node. Pick the checkpointer mode that matches what you actually need. Read the subgraph persistence docs before choosing mode 3.

## The fix

1. No interrupts needed: opt out entirely.

2. Interrupts yes, cross-invocation persistence no: omit the checkpointer (inherited mode). Each invocation gets a unique namespace, parallel execution works, the subgraph starts fresh each time but `interrupt()` still works.

3. Cross-invocation persistence needed: keep `checkpointer=True`. LangGraph assigns position-based namespace suffixes (`calling_node`, `calling_node|1`, ...). For stable names, wrap each subgraph call in a uniquely named node.

## When to use this

- This covers exactly what the title says: MULTIPLE_SUBGRAPHS.
- You are setting this up for the first time, or auditing an existing setup.
- You want the key gotchas in one place before you start.

## When not to use this

- You are doing a different workflow with MULTIPLE_SUBGRAPHS; these steps are specific to the title above.
- You need the full reference docs; this is the short path, not the manual.

## Compatibility

- Not pinned to a specific version; follows current MULTIPLE_SUBGRAPHS behavior.

---

## Vectle citations

- Exact version: <https://vectle.com/skills/skl_6rAjT5zl2pCF492YbD0VyQ/versions/skv_e2YUk00QH9i8XkDXyT29mQ>
- Canonical HTML skill page: <https://vectle.com/skills/skl_6rAjT5zl2pCF492YbD0VyQ>
- Current Markdown URL: <https://vectle.com/skills/skl_6rAjT5zl2pCF492YbD0VyQ/markdown>

## Selected public attribution and original notices/references

```json
{
  "author": "issueatlas",
  "client": null,
  "notices_and_references": [
    "https://docs.langchain.com/oss/python/langgraph/errors/MULTIPLE_SUBGRAPHS"
  ]
}
```
