RCE in Deepwisdom Metagpt
CVE-2026-5970
A vulnerability was detected in FoundationAgents MetaGPT up to 0.8.1. This affects the function check_solution of the component HumanEvalBenchmark/MBPPBenchmark. Performing a manipulation results in code injection. The attack may be initiated remotely. The exploit is now public and may be used. The project was informed of the problem early through a pull request but has not reacted yet.
Published · last modified .
CVSS v3 metric
CVSS v3 base score 7.3 (High). Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:L.
EPSS exploit prediction
EPSS: 0.004 (32.9th percentile), scored .
Very low probability of exploitation in the next 30 days; routine patching cadence is appropriate. 33rd percentile — 32.9% of CVEs in the catalogue have a lower EPSS than this one. How to read EPSS.
EPSS over last 30 days · oldest: 0.007 · newest: 0.004 · change: -0.003
Affected products
- Deepwisdom Metagpt
- Foundationagents Metagpt — versions 0.8.0, 0.8.1
Weakness classification (CWE)
References
- vuldb.com/vuln/356524 (technical-description, VDB Entry, Third Party Advisory, vdb-entry)
- vuldb.com/vuln/356524/cti (signature, Permissions Required, permissions-required, VDB Entry)
- vuldb.com/submit/791693 (Exploit, VDB Entry, Third Party Advisory, third-party-advisory)
- github.com/FoundationAgents/MetaGPT/issues/1942 (issue-tracking, Exploit, exploit, Issue Tracking)
- github.com/FoundationAgents/MetaGPT/pull/1988 (issue-tracking, Patch, patch, Issue Tracking)
- github.com/FoundationAgents/MetaGPT/ (Product, product)
Frequently asked questions
- What is CVE-2026-5970?
- CVE-2026-5970 is a high-severity vulnerability in Deepwisdom Metagpt, classified under Improper Neutralization of Special Elements in Output Used by a Downstream Component (Injection). CVSS score: 7.3/10. Published 2026-04-09.
- How severe is CVE-2026-5970?
- High severity. CVSS v3 base score is 7.3 out of 10.