# LangChain: Path traversal and sandbox escape in LangChain file-search middleware and loaders

Advisory record · `GHSA-gr75-jv2w-4656`

- **Identifiers:** `GHSA-gr75-jv2w-4656` · `CVE-2026-55443` · `PYSEC-2026-2192` · `PYSEC-2026-2556`
- **State:** Confirmed
- **Severity:** CVSS 5.1 Medium
- **Critical:** No
- **Weakness (CWE):** [CWE-22](https://cwe.mitre.org/data/definitions/22.html) · [CWE-59](https://cwe.mitre.org/data/definitions/59.html)
- **Tags:** [Path traversal](https://agentprime.io/advisories/tag/path-traversal.md)
- **Published:** 16 Jun 2026
- **Updated:** 10 Sept 2026
- **Reviewed by a person:** No

Published by the GitHub Advisory Database at 16 Jun 2026, 15:03 UTC. agentprime has not analysed this yet, so whether it affects any app is not known.

## Affected packages

- `langchain` (PyPI): affected `<1.3.9`; fixed in `1.3.9`
- `langchain-anthropic` (PyPI): affected `<1.4.6`; fixed in `1.4.6`

## Machine-written summary

Summary written by Claude Haiku from the source advisory. Not reviewed by a person.

> Path traversal and symlink escape issues in langchain before 1.3.9 and langchain-anthropic before 1.4.6 let an attacker who can influence path values, search patterns, or workspace contents, including through an LLM acting on untrusted input, read files outside the intended root directory. Path-prefix checks that lack a segment boundary can also grant access to sibling paths. Upgrade langchain to 1.3.9 and langchain-anthropic to 1.4.6. Also confine filesystem-backed agent tools to a dedicated directory and do not load configuration from untrusted sources.

Written 1 Nov 2026.

## Not analysed yet

We haven’t published our analysis of this Advisory yet. Once a person has reviewed it, this page will say what’s vulnerable, what has to be true for your app to be affected, and how to check.

## Severity and exploitation

- 5.1 Medium, CVSS 3.1: `CVSS:3.1/AV:L/AC:H/PR:N/UI:N/S:U/C:H/I:N/A:N`, from CNA (GitHub_M)

- **CISA KEV:** Not listed.
- **EPSS:** 0.0021, the 11th percentile, on 16 Nov 2026: FIRST’s estimate of the chance of exploitation in the next 30 days.

## Timeline

- 12 Jun 2026, 16:56 UTC: [Published by the maintainers as a GitHub repository advisory](https://github.com/langchain-ai/langchain/security/advisories/GHSA-gr75-jv2w-4656)
- 31 Oct 2026, 14:11 UTC: First seen by agentprime
- 31 Oct 2026, 14:11 UTC: [Confirmed by OSV](https://osv.dev/vulnerability/GHSA-gr75-jv2w-4656)

## Sources

- [GitHub Advisory Database: GHSA-gr75-jv2w-4656](https://github.com/advisories/GHSA-gr75-jv2w-4656)
- [OSV: GHSA-gr75-jv2w-4656](https://osv.dev/vulnerability/GHSA-gr75-jv2w-4656)
- <https://github.com/langchain-ai/langchain/security/advisories/GHSA-gr75-jv2w-4656>
- <https://nvd.nist.gov/vuln/detail/CVE-2026-55443>
- <https://github.com/langchain-ai/langchain/commit/dcaf7795a3e6590af55c3ff7bda6add6355e9ea6>
- <https://github.com/langchain-ai/langchain>

Facts on this page come from the GitHub Advisory Database (CC-BY 4.0), OSV and the other sources listed, as published. We don’t reproduce full advisory descriptions, which can contain exploit code.

## Check whether this affects you

A Match means an installed version is in the affected range above. Most don’t affect the app; agentprime hasn’t analysed this one yet.

- With the agentprime MCP server (<https://api.agentprime.io/mcp>), ask your agent to check the project: `check_lockfile` sends a lockfile, `check_packages` a list of installed packages, and both return the Matches and the upgrades that clear them. `get_advisory` returns this record. Setup: <https://agentprime.io/mcp.md>
- In a browser: <https://agentprime.io/check/> reads your lockfile in the browser, never uploading it.
- Whether it touches your code, with file and line: the Claude Code plugin, <https://agentprime.io/plugin.md>
- As JSON: <https://api.agentprime.io/v1/advisories/GHSA-gr75-jv2w-4656>, or `POST https://api.agentprime.io/v1/match` with your packages (<https://agentprime.io/api.md>)

---

The web page: <https://agentprime.io/advisories/GHSA-gr75-jv2w-4656/>. Everything for agents: <https://agentprime.io/llms.txt>.
