# ONNX: Out-of-Bounds Read in RNN/GRU/LSTM Shape Inference (Old Opsets, Missing Rank Check)

Advisory record · `GHSA-48m4-24cq-37qf`

- **Identifiers:** `GHSA-48m4-24cq-37qf` · `CVE-2026-115919`
- **State:** Confirmed
- **Severity:** CVSS 6.2 Medium
- **Critical:** No
- **Weakness (CWE):** [CWE-125](https://cwe.mitre.org/data/definitions/125.html)
- **Tags:** [Memory corruption](https://agentprime.io/advisories/tag/memory-corruption.md)
- **Published:** 28 Oct 2026
- **Updated:** 28 Oct 2026
- **Reviewed by a person:** No

Published by the GitHub Advisory Database at 28 Oct 2026, 21:16 UTC. agentprime has not analysed this yet, so whether it affects any app is not known.

## Affected packages

- `onnx` (PyPI): affected `<1.22.0`; fixed in `1.22.0`

## Machine-written summary

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

> The ONNX package before 1.22.0 has an out-of-bounds read in shape inference for old-opset RNN, GRU, and LSTM nodes (opsets 1 to 6) when the first input has a rank other than three. An attacker who can supply a crafted model that passes the ONNX checker and is processed by infer_shapes() can crash the process, causing denial of service; the source reports no confidentiality or integrity impact. Upgrade to ONNX 1.22.0 or later, which validates the input rank before reading dimensions.

Written 31 Oct 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

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

- **CISA KEV:** Not listed.
- **EPSS:** 0.0018, the 7th percentile, on 5 Nov 2026: FIRST’s estimate of the chance of exploitation in the next 30 days.

## Timeline

- 31 Oct 2026, 14:07 UTC: First seen by agentprime
- 31 Oct 2026, 14:07 UTC: [Confirmed by OSV](https://osv.dev/vulnerability/GHSA-48m4-24cq-37qf)

## Sources

- [GitHub Advisory Database: GHSA-48m4-24cq-37qf](https://github.com/advisories/GHSA-48m4-24cq-37qf)
- [OSV: GHSA-48m4-24cq-37qf](https://osv.dev/vulnerability/GHSA-48m4-24cq-37qf)
- <https://github.com/onnx/onnx/security/advisories/GHSA-48m4-24cq-37qf>
- <https://nvd.nist.gov/vuln/detail/CVE-2026-115919>
- <https://github.com/onnx/onnx/pull/7879>
- <https://github.com/onnx/onnx/commit/92e541266af5dd38cfa48aeb92206f4268a066cd>
- <https://github.com/onnx/onnx>
- <https://github.com/onnx/onnx/releases/tag/v1.22.0>

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-48m4-24cq-37qf>, or `POST https://api.agentprime.io/v1/match` with your packages (<https://agentprime.io/api.md>)

---

The web page: <https://agentprime.io/advisories/GHSA-48m4-24cq-37qf/>. Everything for agents: <https://agentprime.io/llms.txt>.