{"id":"CVE-2026-40068","aliases":["EUVD-2026-27502","GHSA-q5hj-mxqh-vv77"],"status":"published","title":"Claude Code arbitrary code execution via git worktree commondir trust dialog bypass","description":"In versions 2.1.63 through 2.1.83 of Claude Code, the folder trust determination logic used the git worktree commondir file without validating its contents. An attacker could craft a malicious repository with a commondir file pointing to a path the victim had previously trusted, causing Claude Code to bypass its trust confirmation dialog and immediately execute hooks defined in `.claude/settings.json`. Exploitation requires the victim to clone the malicious repository and run Claude Code within it, and the attacker must know or guess a path the victim had already trusted. This issue has been fixed in version 2.1.84.","severity":{"version":"4.0","score":7.7,"vector":"CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:P/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N","level":"high","assessor":"cna","from":"cve"},"epss":{"day":"2026-11-11","score":0.00314,"percentile":0.24635,"model":"v2026.06.15"},"facts":{"in_kev":false,"known_exploited":false,"public_exploit":false,"fix_available":true},"cwes":["CWE-20","CWE-77"],"published":"2026-04-24T16:34:03Z","modified":"2026-10-28T02:31:15Z","fixes":[{"ecosystem":"npm","package":"@anthropic-ai/claude-code","fixed_in":["2.1.84"],"sources":["ghsa","osv"]}],"affected":[{"source":"ghsa","ecosystem":"npm","package":"@anthropic-ai/claude-code","ranges":[{"type":"ECOSYSTEM","events":[{"introduced":"2.1.63"},{"fixed":"2.1.84"}]}]},{"source":"osv","ecosystem":"npm","package":"@anthropic-ai/claude-code","ranges":[{"type":"SEMVER","events":[{"introduced":"2.1.63"},{"fixed":"2.1.84"}]}]}],"products":[{"vendor":"anthropic","product":"claude_code"},{"vendor":"anthropics","product":"claude-code"}],"references":[{"url":"https://github.com/anthropics/claude-code/security/advisories/GHSA-q5hj-mxqh-vv77","tags":["x_refsource_CONFIRM","Vendor Advisory","advisory","web"]},{"url":"https://nvd.nist.gov/vuln/detail/CVE-2026-40068","tags":["advisory"]},{"url":"https://github.com/advisories/GHSA-q5hj-mxqh-vv77"},{"url":"https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/40xxx/CVE-2026-40068.json","tags":["advisory"]},{"url":"https://github.com/anthropics/claude-code","tags":["package"]}],"sources":{"cve":[{"id":"CVE-2026-40068","title":"Claude Code arbitrary code execution via git worktree commondir trust dialog bypass","description":"In versions 2.1.63 through 2.1.83 of Claude Code, the folder trust determination logic used the git worktree commondir file without validating its contents. An attacker could craft a malicious repository with a commondir file pointing to a path the victim had previously trusted, causing Claude Code to bypass its trust confirmation dialog and immediately execute hooks defined in `.claude/settings.json`. Exploitation requires the victim to clone the malicious repository and run Claude Code within it, and the attacker must know or guess a path the victim had already trusted. This issue has been fixed in version 2.1.84.","status":"published","published":"2026-05-05T20:52:26.089Z","modified":"2026-05-06T15:24:52.454Z","cvss":[{"version":"4.0","score":7.7,"vector":"CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:P/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N","level":"high","assessor":"cna"}],"cwes":["CWE-20","CWE-77"],"updated_at":"2026-11-02T13:43:22.599321Z"}],"euvd":[{"id":"EUVD-2026-27502","description":"In versions 2.1.63 through 2.1.83 of Claude Code, the folder trust determination logic used the git worktree commondir file without validating its contents. An attacker could craft a malicious repository with a commondir file pointing to a path the victim had previously trusted, causing Claude Code to bypass its trust confirmation dialog and immediately execute hooks defined in `.claude/settings.json`. Exploitation requires the victim to clone the malicious repository and run Claude Code within it, and the attacker must know or guess a path the victim had already trusted. This issue has been fixed in version 2.1.84.","status":"published","published":"2026-05-05T20:52:26Z","modified":"2026-05-06T15:24:52Z","cvss":[{"version":"4.0","score":7.7,"vector":"CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:P/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N","level":"high"}],"updated_at":"2026-11-02T19:05:16.191098Z"}],"ghsa":[{"id":"GHSA-q5hj-mxqh-vv77","title":"Claude Code: Trust Dialog Bypass via Git Worktree Spoofing Allows Arbitrary Code Execution","description":"Claude Code used the git worktree `commondir` file when determining folder trust but did not validate its contents. By crafting a repository with a `commondir` file pointing to a path the victim had previously trusted, an attacker could bypass the trust dialog and immediately execute malicious hooks defined in `.claude/settings.json`. Exploiting this required the victim to clone a malicious repository and run Claude Code within it, and for the attacker to know or guess a path the victim had already trusted.\n\nUsers on standard Claude Code auto-update have received this fix already. Users performing manual updates are advised to update to the latest version.\n\nClaude Code thanks [hackerone.com/masato_anzai](https://hackerone.com/masato_anzai) for reporting this issue.","status":"published","published":"2026-04-24T16:34:03Z","modified":"2026-05-08T15:31:39Z","cvss":[{"version":"4.0","score":7.7,"vector":"CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:P/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N","level":"high"}],"cwes":["CWE-20","CWE-77"],"updated_at":"2026-11-02T14:40:03.762927Z"}],"nvd":[{"id":"CVE-2026-40068","description":"In versions 2.1.63 through 2.1.83 of Claude Code, the folder trust determination logic used the git worktree commondir file without validating its contents. An attacker could craft a malicious repository with a commondir file pointing to a path the victim had previously trusted, causing Claude Code to bypass its trust confirmation dialog and immediately execute hooks defined in `.claude/settings.json`. Exploitation requires the victim to clone the malicious repository and run Claude Code within it, and the attacker must know or guess a path the victim had already trusted. This issue has been fixed in version 2.1.84.","status":"published","published":"2026-05-05T21:16:23.093Z","modified":"2026-07-24T21:10:00.143Z","cvss":[{"version":"4.0","score":7.7,"vector":"CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:P/VC:H/VI:H/VA:H/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":"high","assessor":"secondary"},{"version":"3.1","score":8.8,"vector":"CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H","level":"high","assessor":"primary"}],"cwes":["CWE-20","CWE-77"],"updated_at":"2026-11-02T14:12:17.418622Z"}],"osv":[{"id":"CVE-2026-40068","title":"Claude Code arbitrary code execution via git worktree commondir trust dialog bypass","description":"In versions 2.1.63 through 2.1.83 of Claude Code, the folder trust determination logic used the git worktree commondir file without validating its contents. An attacker could craft a malicious repository with a commondir file pointing to a path the victim had previously trusted, causing Claude Code to bypass its trust confirmation dialog and immediately execute hooks defined in `.claude/settings.json`. Exploitation requires the victim to clone the malicious repository and run Claude Code within it, and the attacker must know or guess a path the victim had already trusted. This issue has been fixed in version 2.1.84.","status":"published","published":"2026-05-05T20:52:26.089Z","modified":"2026-10-28T02:31:15.102869245Z","cvss":[{"version":"4.0","score":7.7,"vector":"CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:P/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N","level":"high"}],"cwes":["CWE-20","CWE-77"],"updated_at":"2026-11-02T15:45:03.040464Z"},{"id":"GHSA-q5hj-mxqh-vv77","title":"Claude Code: Trust Dialog Bypass via Git Worktree Spoofing Allows Arbitrary Code Execution","description":"Claude Code used the git worktree `commondir` file when determining folder trust but did not validate its contents. By crafting a repository with a `commondir` file pointing to a path the victim had previously trusted, an attacker could bypass the trust dialog and immediately execute malicious hooks defined in `.claude/settings.json`. Exploiting this required the victim to clone a malicious repository and run Claude Code within it, and for the attacker to know or guess a path the victim had already trusted.\n\nUsers on standard Claude Code auto-update have received this fix already. Users performing manual updates are advised to update to the latest version.\n\nClaude Code thanks [hackerone.com/masato_anzai](https://hackerone.com/masato_anzai) for reporting this issue.","status":"published","published":"2026-04-24T16:34:03Z","modified":"2026-05-08T15:50:36.976042Z","cvss":[{"version":"4.0","score":7.7,"vector":"CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:P/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N","level":"high"}],"cwes":["CWE-20","CWE-77"],"updated_at":"2026-11-02T15:46:54.055966Z"}]}}