{"action":"Upgrade praisonai to 4.6.40 or later.","affected":[{"fixed":"4.6.40","introduced":"0"}],"component":"praisonai","cve_ids":["CVE-2026-47391"],"ecosystem":"PyPI","epss":0.01498,"exploited":false,"fixed_versions":["4.6.40"],"ghsa_id":"GHSA-vg22-4gmj-prxw","id":"tcs_01M5KHBKRJ1XZ629S2MR1S7JC8","modified_at":"2026-07-13T16:15:33.187039504Z","published_at":"2026-06-29T11:50:48.861322Z","references":["https://osv.dev/vulnerability/PYSEC-2026-478","https://github.com/MervinPraison/PraisonAI/security/advisories/GHSA-vg22-4gmj-prxw","https://github.com/MervinPraison/PraisonAI","https://pypi.org/project/praisonai","https://github.com/advisories/GHSA-vg22-4gmj-prxw","https://nvd.nist.gov/vuln/detail/CVE-2026-47391"],"severity":{"label":"critical","score":9.8},"sources":[{"source":"osv","source_id":"PYSEC-2026-478","url":"https://osv.dev/vulnerability/PYSEC-2026-478"}],"summary":"PraisonAI's unauthenticated A2A official example can reach real LLM-driven `eval()` tool execution"}