hf_doc_fetch
on https://huggingface.co/mcp
Severity
1 finding on this tool
- mediumindirect injectionf-r6-hf_doc_fetch
Tool `hf_doc_fetch` fetches external web content -- indirect-injection surface
Description: "Fetch a document from the Hugging Face or Gradio documentation library. For large documents, use offset to get subsequent chunks." -- this tool pulls externally-controlled content into the agent's context window, the canonical indirect-injection vector. Even when the user supplies the URL, content at that URL can carry hostile instructions.
fix: Sandbox the fetched content: strip prompts before forwarding to the model, constrain to an allow-list of domains, and route through capframe-guard with a `domain in [...]` caveat.
OWASP LLM01NIST MEASURE-2.3ATLAS T0051
About this tool
hf_doc_fetch is one of 8 tools exposed by Hugging Face Hub MCP. The server scored 78/100 overall against the capframe rule engine (source: http). Last scanned 2026-06-05.
The findings above are emitted by the public capframe.findings.v1 schema. Disagree with one? Open an issue.