BREW-LANGGRAPH-CLI-CVE-2… (GHSA-2G6R-C272-W58R)
Vulnerability from osv_homebrew – Published: 2026-08-13 17:02 – Updated: 2026-09-10 00:18 – Source websiteServer-Side Request Forgery (SSRF) in ChatOpenAI Image Token Counting
Summary
The ChatOpenAI.get_num_tokens_from_messages() method fetches arbitrary image_url values without validation when computing token counts for vision-enabled models. This allows attackers to trigger Server-Side Request Forgery (SSRF) attacks by providing malicious image URLs in user input.
Severity
Low - The vulnerability allows SSRF attacks but has limited impact due to: - Responses are not returned to the attacker (blind SSRF) - Default 5-second timeout limits resource exhaustion - Non-image responses fail at PIL image parsing
Impact
An attacker who can control image URLs passed to get_num_tokens_from_messages() can:
- Trigger HTTP requests from the application server to arbitrary internal or external URLs
- Cause the server to access internal network resources (private IPs, cloud metadata endpoints)
- Cause minor resource consumption through image downloads (bounded by timeout)
Note: This vulnerability occurs during token counting, which may happen outside of model invocation (e.g., in logging, metrics, or token budgeting flows).
Details
The vulnerable code path:
1. get_num_tokens_from_messages() processes messages containing image_url content blocks
2. For images without detail: "low", it calls _url_to_size() to fetch the image and compute token counts
3. _url_to_size() performs httpx.get(image_source) on any URL without validation
4. Prior to the patch, there was no SSRF protection, size limits, or explicit timeout
File: libs/partners/openai/langchain_openai/chat_models/base.py
Patches
The vulnerability has been patched in langchain-openai==1.1.9 (requires langchain-core==1.2.11).
The patch adds:
1. SSRF validation using langchain_core._security._ssrf_protection.validate_safe_url() to block:
- Private IP ranges (RFC 1918, loopback, link-local)
- Cloud metadata endpoints (169.254.169.254, etc.)
- Invalid URL schemes
2. Explicit size limits (50 MB maximum, matching OpenAI's payload limit)
3. Explicit timeout (5 seconds, same as httpx.get default)
4. Allow disabling image fetching via allow_fetching_images=False parameter
Workarounds
If you cannot upgrade immediately:
- Sanitize input: Validate and filter
image_urlvalues before passing messages to token counting or model invocation - Use network controls: Implement egress filtering to prevent outbound requests to private IPs
{
"affected": [
{
"ecosystem_specific": {
"fix": "bump",
"range_state": "fixed",
"resource": "langchain-core",
"resource_purl": "pkg:pypi/langchain-core@1.4.9",
"upstream_fixed_in": "1.2.11"
},
"package": {
"ecosystem": "Homebrew",
"name": "langgraph-cli",
"purl": "pkg:brew/langgraph-cli"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "0.4.31"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"database_specific": {
"confidence": "high",
"source": "matched",
"strategy": "registry",
"upstream_evidence": [
{
"ecosystem": "PyPI",
"key": "pkg:pypi/langchain-core@1.4.9",
"name": "langchain-core",
"resource": "langchain-core",
"strategy": "registry",
"subject_version": "1.4.9"
}
]
},
"details": "## Server-Side Request Forgery (SSRF) in ChatOpenAI Image Token Counting\n\n### Summary\nThe `ChatOpenAI.get_num_tokens_from_messages()` method fetches arbitrary `image_url` values without validation when computing token counts for vision-enabled models. This allows attackers to trigger Server-Side Request Forgery (SSRF) attacks by providing malicious image URLs in user input.\n\n### Severity\n**Low** - The vulnerability allows SSRF attacks but has limited impact due to:\n- Responses are not returned to the attacker (blind SSRF)\n- Default 5-second timeout limits resource exhaustion\n- Non-image responses fail at PIL image parsing\n\n### Impact\nAn attacker who can control image URLs passed to `get_num_tokens_from_messages()` can:\n- Trigger HTTP requests from the application server to arbitrary internal or external URLs\n- Cause the server to access internal network resources (private IPs, cloud metadata endpoints)\n- Cause minor resource consumption through image downloads (bounded by timeout)\n\n**Note:** This vulnerability occurs during token counting, which may happen outside of model invocation (e.g., in logging, metrics, or token budgeting flows).\n\n### Details\nThe vulnerable code path:\n1. `get_num_tokens_from_messages()` processes messages containing `image_url` content blocks\n2. For images without `detail: \"low\"`, it calls `_url_to_size()` to fetch the image and compute token counts\n3. `_url_to_size()` performs `httpx.get(image_source)` on any URL without validation\n4. Prior to the patch, there was no SSRF protection, size limits, or explicit timeout\n\n**File:** `libs/partners/openai/langchain_openai/chat_models/base.py`\n\n### Patches\nThe vulnerability has been patched in `langchain-openai==1.1.9` (requires `langchain-core==1.2.11`).\n\nThe patch adds:\n1. **SSRF validation** using `langchain_core._security._ssrf_protection.validate_safe_url()` to block:\n - Private IP ranges (RFC 1918, loopback, link-local)\n - Cloud metadata endpoints (169.254.169.254, etc.)\n - Invalid URL schemes\n2. **Explicit size limits** (50 MB maximum, matching OpenAI\u0027s payload limit)\n3. **Explicit timeout** (5 seconds, same as `httpx.get` default)\n4. **Allow disabling image fetching** via `allow_fetching_images=False` parameter\n\n### Workarounds\nIf you cannot upgrade immediately:\n\n1. **Sanitize input:** Validate and filter `image_url` values before passing messages to token counting or model invocation\n2. **Use network controls:** Implement egress filtering to prevent outbound requests to private IPs",
"id": "BREW-langgraph-cli-CVE-2026-26013",
"modified": "2026-09-10T00:18:25Z",
"published": "2026-08-13T17:02:14Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/langchain-ai/langchain/security/advisories/GHSA-2g6r-c272-w58r"
},
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-26013"
},
{
"type": "WEB",
"url": "https://github.com/langchain-ai/langchain/commit/2b4b1dc29a833d4053deba4c2b77a3848c834565"
},
{
"type": "PACKAGE",
"url": "https://github.com/langchain-ai/langchain"
},
{
"type": "WEB",
"url": "https://github.com/langchain-ai/langchain/releases/tag/langchain-core%3D%3D1.2.11"
}
],
"schema_version": "1.7.3",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:N/A:L",
"type": "CVSS_V3"
}
],
"summary": "LangChain affected by SSRF via image_url token counting in ChatOpenAI.get_num_tokens_from_messages",
"upstream": [
"GHSA-2g6r-c272-w58r",
"CVE-2026-26013",
"PYSEC-2026-2562"
]
}
Sightings
| Author | Source | Type | Date | Other |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or observed by the user.
- Confirmed: The vulnerability has been validated from an analyst's perspective.
- Published Proof of Concept: A public proof of concept is available for this vulnerability.
- Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
- Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
- Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
- Not confirmed: The user expressed doubt about the validity of the vulnerability.
- Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.
Browse all ATT&CK techniques and the vulnerabilities related to each.
Related by attack behaviour
Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.