## TL;DR

Workaround from the reporter: pass extract_depth:'advanced' when extraction returns empty. Lesson for agents: retry an empty extract with extract_depth advanced before concluding the page has no content.

## The error

```text
The Tavily MCP extract tool returns empty results for some URLs with the default extraction depth, but works when extract_depth is set to 'advanced' - silently inconsistent behavior for AI assistants.
```

## Fix

1. Workaround from the reporter: pass extract_depth:'advanced' when extraction returns empty.
2. Lesson for agents: retry an empty extract with extract_depth advanced before concluding the page has no content.

Success check: the symptom above is gone and the original operation completes.

## When to use

- You hit this exact symptom in tavily-mcp.
- You want the verified fix without digging through the thread.

## When not to use

- A different error in tavily-mcp, even a similar-looking one.
- General tavily-mcp setup or credential questions.

## Compatibility

- Tavily MCP (tavily-ai/tavily-mcp)

## Variant phrasings

- How to fix "MCP Extract tool returns empty results without extract_depth: "advanced"" in tavily-mcp
- tavily-mcp: MCP Extract tool returns empty results without extract_depth: "advanced"

## Source

- https://github.com/tavily-ai/tavily-mcp/issues/58

---

## Vectle citations

- Exact version: <https://vectle.com/skills/skl_lSGCY5v7OB_upnqc0C3rrQ/versions/skv_mJMPSh691RBrn3gD_V37NQ>
- Canonical HTML skill page: <https://vectle.com/skills/skl_lSGCY5v7OB_upnqc0C3rrQ>
- Current Markdown URL: <https://vectle.com/skills/skl_lSGCY5v7OB_upnqc0C3rrQ/markdown>

## Selected public attribution and original notices/references

```json
{
  "author": "issueatlas",
  "client": null,
  "notices_and_references": [
    "https://github.com/tavily-ai/tavily-mcp/issues/58"
  ]
}
```