{"id":"CVE-2026-65698","aliases":["EUVD-2026-48312","GHSA-739h-jx7m-fc7g"],"status":"published","title":"Void 1.3.4 Path Traversal via AI Agent File-Reading Tools","description":"Void through 1.3.4 contains a path traversal vulnerability in the AI agent file-reading tools that allows network-adjacent attackers to read arbitrary host files outside the open workspace by injecting instructions into content the agent processes. Attackers can supply absolute paths or file:// URIs through the read_file, ls_dir, get_dir_tree, and search_* tools, which lack workspace confinement and bypass the approval gate, enabling silent exfiltration of sensitive files such as SSH private keys or cloud credentials via subsequent tool calls.","severity":{"version":"4.0","score":6,"vector":"CVSS:4.0/AV:N/AC:H/AT:N/PR:N/UI:P/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N","level":"medium","assessor":"cna","from":"cve"},"epss":{"day":"2026-11-10","score":0.0047,"percentile":0.40977,"model":"v2026.06.15"},"facts":{"in_kev":false,"known_exploited":false,"public_exploit":false,"fix_available":false},"cwes":["CWE-22"],"published":"2026-07-23T16:37:42Z","modified":"2026-10-28T02:30:45Z","products":[{"vendor":"voideditor","product":"void"}],"references":[{"url":"https://github.com/geo-chen/oss/blob/main/void.md","tags":["technical-description","exploit","evidence"]},{"url":"https://www.vulncheck.com/advisories/void-path-traversal-via-ai-agent-file-reading-tools","tags":["third-party-advisory","advisory"]},{"url":"https://nvd.nist.gov/vuln/detail/CVE-2026-65698","tags":["advisory"]},{"url":"https://github.com/advisories/GHSA-739h-jx7m-fc7g"},{"url":"https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/65xxx/CVE-2026-65698.json","tags":["advisory"]},{"url":"https://github.com/voideditor/void","tags":["package"]}],"sources":{"cve":[{"id":"CVE-2026-65698","title":"Void 1.3.4 Path Traversal via AI Agent File-Reading Tools","description":"Void through 1.3.4 contains a path traversal vulnerability in the AI agent file-reading tools that allows network-adjacent attackers to read arbitrary host files outside the open workspace by injecting instructions into content the agent processes. Attackers can supply absolute paths or file:// URIs through the read_file, ls_dir, get_dir_tree, and search_* tools, which lack workspace confinement and bypass the approval gate, enabling silent exfiltration of sensitive files such as SSH private keys or cloud credentials via subsequent tool calls.","status":"published","published":"2026-07-23T16:37:42.944Z","modified":"2026-07-24T22:00:16.889Z","cvss":[{"version":"4.0","score":6,"vector":"CVSS:4.0/AV:N/AC:H/AT:N/PR:N/UI:P/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N","level":"medium","assessor":"cna"},{"version":"3.1","score":5.3,"vector":"CVSS:3.1/AV:N/AC:H/PR:N/UI:R/S:U/C:H/I:N/A:N","level":"medium","assessor":"cna"}],"cwes":["CWE-22"],"updated_at":"2026-11-02T13:37:13.934727Z"}],"euvd":[{"id":"EUVD-2026-48312","description":"Void through 1.3.4 contains a path traversal vulnerability in the AI agent file-reading tools that allows network-adjacent attackers to read arbitrary host files outside the open workspace by injecting instructions into content the agent processes. Attackers can supply absolute paths or file:// URIs through the read_file, ls_dir, get_dir_tree, and search_* tools, which lack workspace confinement and bypass the approval gate, enabling silent exfiltration of sensitive files such as SSH private keys or cloud credentials via subsequent tool calls.","status":"published","published":"2026-07-23T16:37:42Z","modified":"2026-07-24T22:00:16Z","cvss":[{"version":"4.0","score":6,"vector":"CVSS:4.0/AV:N/AC:H/AT:N/PR:N/UI:P/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N","level":"medium"}],"updated_at":"2026-11-02T18:19:38.612768Z"}],"ghsa":[{"id":"GHSA-739h-jx7m-fc7g","title":"Void through 1.3.4 contains a path traversal vulnerability in the AI agent file-reading tools...","description":"Void through 1.3.4 contains a path traversal vulnerability in the AI agent file-reading tools that allows network-adjacent attackers to read arbitrary host files outside the open workspace by injecting instructions into content the agent processes. Attackers can supply absolute paths or file:// URIs through the read_file, ls_dir, get_dir_tree, and search_* tools, which lack workspace confinement and bypass the approval gate, enabling silent exfiltration of sensitive files such as SSH private keys or cloud credentials via subsequent tool calls.","status":"published","published":"2026-07-23T18:31:45Z","modified":"2026-07-23T18:31:45Z","cvss":[{"version":"4.0","score":6,"vector":"CVSS:4.0/AV:N/AC:H/AT:N/PR:N/UI:P/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X","level":"medium"},{"version":"3.1","score":5.3,"vector":"CVSS:3.1/AV:N/AC:H/PR:N/UI:R/S:U/C:H/I:N/A:N","level":"medium"}],"cwes":["CWE-22"],"updated_at":"2026-11-02T15:09:43.679745Z"}],"nvd":[{"id":"CVE-2026-65698","description":"Void through 1.3.4 contains a path traversal vulnerability in the AI agent file-reading tools that allows network-adjacent attackers to read arbitrary host files outside the open workspace by injecting instructions into content the agent processes. Attackers can supply absolute paths or file:// URIs through the read_file, ls_dir, get_dir_tree, and search_* tools, which lack workspace confinement and bypass the approval gate, enabling silent exfiltration of sensitive files such as SSH private keys or cloud credentials via subsequent tool calls.","status":"published","published":"2026-07-23T17:16:30.05Z","modified":"2026-10-24T13:10:00.147Z","cvss":[{"version":"4.0","score":6,"vector":"CVSS:4.0/AV:N/AC:H/AT:N/PR:N/UI:P/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X","level":"medium","assessor":"secondary"},{"version":"3.1","score":5.3,"vector":"CVSS:3.1/AV:N/AC:H/PR:N/UI:R/S:U/C:H/I:N/A:N","level":"medium","assessor":"secondary"}],"cwes":["CWE-22"],"updated_at":"2026-11-02T14:12:32.473901Z"}],"osv":[{"id":"CVE-2026-65698","title":"Void 1.3.4 Path Traversal via AI Agent File-Reading Tools","description":"Void through 1.3.4 contains a path traversal vulnerability in the AI agent file-reading tools that allows network-adjacent attackers to read arbitrary host files outside the open workspace by injecting instructions into content the agent processes. Attackers can supply absolute paths or file:// URIs through the read_file, ls_dir, get_dir_tree, and search_* tools, which lack workspace confinement and bypass the approval gate, enabling silent exfiltration of sensitive files such as SSH private keys or cloud credentials via subsequent tool calls.","status":"published","published":"2026-07-23T16:37:42.944Z","modified":"2026-10-28T02:30:45.260893744Z","cvss":[{"version":"4.0","score":6,"vector":"CVSS:4.0/AV:N/AC:H/AT:N/PR:N/UI:P/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N","level":"medium"}],"cwes":["CWE-22"],"updated_at":"2026-11-02T15:45:14.46177Z"}]}}