Search
Find a vulnerability
Search criteria
ⓘ
Use this form to refine search results.
Full-text search supports keyword queries with ranking and filtering.
You can combine vendor, product, and sources to narrow results.
Enable “Apply ordering” to sort by date instead of relevance.
72 vulnerabilities by Anthropic
GCVE-1337-2026-000000000… (CVE-2026-7574)
Vulnerability from gna-1337 – Published: 2026-06-23 23:54 – Updated: 2026-06-23 23:54
VLAI
Title
Anthropic Claude Desktop Cowork VM Image Contents Not Validated Before Use
Summary
Anthropic Claude Desktop Cowork VM image handling (confirmed across v1.1348.0 through v1.2278.0, including v1.1348.0, v1.1617.0, and v1.2278.0) validates only file presence and a version marker string before booting rootfs.img, but does not verify image content integrity at time-of-use. A local attacker with unprivileged code execution as the victim macOS user can modify the VM root filesystem image and have it trusted on subsequent Cowork VM boots, enabling persistent arbitrary code execution in the VM and access to host-mounted directories. The estimated CWE mapping is CWE-353 (Missing Support for Integrity Check).
Severity
8.7 (High)
CWE
- CWE-353 - Missing support for integrity check
References
2 references
| URL | Tags |
|---|---|
| https://cfp.recon.cx/recon-2026/talk/DZUQYU/ | related |
| https://takeonme.org/gcves/GCVE-1337-2026-0000000… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| Anthropic | Claude Desktop Cowork |
Affected:
1.1348.0 , ≤ 1.2278.0
(semver)
|
Credits
{
"containers": {
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"product": "Claude Desktop Cowork",
"vendor": "Anthropic",
"versions": [
{
"lessThanOrEqual": "1.2278.0",
"status": "affected",
"version": "1.1348.0",
"versionType": "semver"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "finder",
"value": "Todd Manning"
},
{
"lang": "en",
"type": "coordinator",
"value": "todb of AHA!"
}
],
"descriptions": [
{
"lang": "en",
"supportingMedia": [
{
"base64": false,
"type": "text/html",
"value": "Anthropic Claude Desktop Cowork VM image handling (confirmed across v1.1348.0 through v1.2278.0, including v1.1348.0, v1.1617.0, and v1.2278.0) validates only file presence and a version marker string before booting rootfs.img, but does not verify image content integrity at time-of-use. A local attacker with unprivileged code execution as the victim macOS user can modify the VM root filesystem image and have it trusted on subsequent Cowork VM boots, enabling persistent arbitrary code execution in the VM and access to host-mounted directories. The estimated CWE mapping is CWE-353 (Missing Support for Integrity Check)."
}
],
"value": "Anthropic Claude Desktop Cowork VM image handling (confirmed across v1.1348.0 through v1.2278.0, including v1.1348.0, v1.1617.0, and v1.2278.0) validates only file presence and a version marker string before booting rootfs.img, but does not verify image content integrity at time-of-use. A local attacker with unprivileged code execution as the victim macOS user can modify the VM root filesystem image and have it trusted on subsequent Cowork VM boots, enabling persistent arbitrary code execution in the VM and access to host-mounted directories. The estimated CWE mapping is CWE-353 (Missing Support for Integrity Check)."
}
],
"metrics": [
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "LOCAL",
"availabilityImpact": "LOW",
"baseScore": 8.7,
"baseSeverity": "HIGH",
"confidentialityImpact": "HIGH",
"integrityImpact": "HIGH",
"privilegesRequired": "LOW",
"scope": "CHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:L",
"version": "3.1"
},
"format": "CVSS",
"scenarios": [
{
"lang": "en",
"value": "GENERAL"
}
]
},
{
"other": {
"content": {
"options": [
{
"Exploitation": "poc"
},
{
"Automatable": "no"
},
{
"Technical Impact": "total"
}
],
"role": "CNA",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-353",
"description": "CWE-353 Missing support for integrity check",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-06-23T23:54:21.635Z",
"orgId": "26969f82-7e87-44d8-9cb5-f6fb926ddd43",
"shortName": "AHA"
},
"references": [
{
"tags": [
"related"
],
"url": "https://cfp.recon.cx/recon-2026/talk/DZUQYU/"
},
{
"tags": [
"third-party-advisory"
],
"url": "https://takeonme.org/gcves/GCVE-1337-2026-00000000000000000000000000000000000000000000000001110111111101011111111110000000000000000000000000000000000000000000000000000001010"
}
],
"source": {
"discovery": "EXTERNAL"
},
"title": "Anthropic Claude Desktop Cowork VM Image Contents Not Validated Before Use",
"x_generator": {
"engine": "Vulnogram 1.0.2"
}
}
},
"cveMetadata": {
"assignerOrgId": "26969f82-7e87-44d8-9cb5-f6fb926ddd43",
"assignerShortName": "AHA",
"cveId": "CVE-2026-7574",
"datePublished": "2026-06-23T23:54:21.635Z",
"dateReserved": "2026-05-01T02:06:20.230Z",
"dateUpdated": "2026-06-23T23:54:21.635Z",
"serial": 1,
"state": "PUBLISHED",
"vulnId": "GCVE-1337-2026-00000000000000000000000000000000000000000000000001110111111101011111111110000000000000000000000000000000000000000000000000000001010"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-7574 (GCVE-0-2026-7574)
Vulnerability from cvelistv5 – Published: 2026-06-23 23:54 – Updated: 2026-06-24 13:04
VLAI
Title
Anthropic Claude Desktop Cowork VM Image Contents Not Validated Before Use
Summary
Anthropic Claude Desktop Cowork VM image handling (confirmed across v1.1348.0 through v1.2278.0, including v1.1348.0, v1.1617.0, and v1.2278.0) validates only file presence and a version marker string before booting rootfs.img, but does not verify image content integrity at time-of-use. A local attacker with unprivileged code execution as the victim macOS user can modify the VM root filesystem image and have it trusted on subsequent Cowork VM boots, enabling persistent arbitrary code execution in the VM and access to host-mounted directories. The estimated CWE mapping is CWE-353 (Missing Support for Integrity Check).
Severity
8.7 (High)
SSVC
Exploitation: none
Automatable: no
Technical Impact: total
CISA Coordinator (v2.0.3)
CWE
- CWE-353 - Missing support for integrity check
Assigner
References
2 references
| URL | Tags |
|---|---|
| https://cfp.recon.cx/recon-2026/talk/DZUQYU/ | related |
| https://takeonme.org/gcves/GCVE-1337-2026-0000000… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| Anthropic | Claude Desktop Cowork |
Affected:
1.1348.0 , ≤ 1.2278.0
(semver)
|
Credits
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-7574",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "total"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-06-24T13:03:53.789182Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-06-24T13:04:11.984Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"product": "Claude Desktop Cowork",
"vendor": "Anthropic",
"versions": [
{
"lessThanOrEqual": "1.2278.0",
"status": "affected",
"version": "1.1348.0",
"versionType": "semver"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "finder",
"value": "Todd Manning"
},
{
"lang": "en",
"type": "coordinator",
"value": "todb of AHA!"
}
],
"descriptions": [
{
"lang": "en",
"supportingMedia": [
{
"base64": false,
"type": "text/html",
"value": "Anthropic Claude Desktop Cowork VM image handling (confirmed across v1.1348.0 through v1.2278.0, including v1.1348.0, v1.1617.0, and v1.2278.0) validates only file presence and a version marker string before booting rootfs.img, but does not verify image content integrity at time-of-use. A local attacker with unprivileged code execution as the victim macOS user can modify the VM root filesystem image and have it trusted on subsequent Cowork VM boots, enabling persistent arbitrary code execution in the VM and access to host-mounted directories. The estimated CWE mapping is CWE-353 (Missing Support for Integrity Check)."
}
],
"value": "Anthropic Claude Desktop Cowork VM image handling (confirmed across v1.1348.0 through v1.2278.0, including v1.1348.0, v1.1617.0, and v1.2278.0) validates only file presence and a version marker string before booting rootfs.img, but does not verify image content integrity at time-of-use. A local attacker with unprivileged code execution as the victim macOS user can modify the VM root filesystem image and have it trusted on subsequent Cowork VM boots, enabling persistent arbitrary code execution in the VM and access to host-mounted directories. The estimated CWE mapping is CWE-353 (Missing Support for Integrity Check)."
}
],
"metrics": [
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "LOCAL",
"availabilityImpact": "LOW",
"baseScore": 8.7,
"baseSeverity": "HIGH",
"confidentialityImpact": "HIGH",
"integrityImpact": "HIGH",
"privilegesRequired": "LOW",
"scope": "CHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:L",
"version": "3.1"
},
"format": "CVSS",
"scenarios": [
{
"lang": "en",
"value": "GENERAL"
}
]
},
{
"other": {
"content": {
"options": [
{
"Exploitation": "poc"
},
{
"Automatable": "no"
},
{
"Technical Impact": "total"
}
],
"role": "CNA",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-353",
"description": "CWE-353 Missing support for integrity check",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-06-23T23:54:21.635Z",
"orgId": "26969f82-7e87-44d8-9cb5-f6fb926ddd43",
"shortName": "AHA"
},
"references": [
{
"tags": [
"related"
],
"url": "https://cfp.recon.cx/recon-2026/talk/DZUQYU/"
},
{
"tags": [
"third-party-advisory"
],
"url": "https://takeonme.org/gcves/GCVE-1337-2026-00000000000000000000000000000000000000000000000001110111111101011111111110000000000000000000000000000000000000000000000000000001010"
}
],
"source": {
"discovery": "EXTERNAL"
},
"title": "Anthropic Claude Desktop Cowork VM Image Contents Not Validated Before Use",
"x_generator": {
"engine": "Vulnogram 1.0.2"
}
}
},
"cveMetadata": {
"assignerOrgId": "26969f82-7e87-44d8-9cb5-f6fb926ddd43",
"assignerShortName": "AHA",
"cveId": "CVE-2026-7574",
"datePublished": "2026-06-23T23:54:21.635Z",
"dateReserved": "2026-05-01T02:06:20.230Z",
"dateUpdated": "2026-06-24T13:04:11.984Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-44470 (GCVE-0-2026-44470)
Vulnerability from cvelistv5 – Published: 2026-05-13 15:41 – Updated: 2026-05-13 18:37
VLAI
Title
Claude Desktop: Local Privilege Escalation via Directory Junction in CoworkVMService
Summary
The Claude Desktop app gives you Claude Code with a graphical interface built for running multiple sessions side by side. Prior to 1.3834.0, the CoworkVMService component in Claude Desktop for Windows ran as SYSTEM and did not validate whether the VM bundle directory was a real directory or an NTFS directory junction before creating files within it. A local non-elevated user could replace the user-writable VM bundle directory with a directory junction pointing to an attacker-chosen location, causing the service to create a SYSTEM-owned file in an arbitrary directory. This could be leveraged for local privilege escalation. This vulnerability is fixed in 1.3834.0.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: total
CISA Coordinator (v2.0.3)
CWE
Assigner
References
1 reference
| URL | Tags |
|---|---|
| https://github.com/anthropics/claude-code/securit… | x_refsource_CONFIRM |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| anthropics | claude-code |
Affected:
< 1.3834.0
|
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-44470",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "total"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-05-13T18:36:56.249906Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-05-13T18:37:19.139Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"product": "claude-code",
"vendor": "anthropics",
"versions": [
{
"status": "affected",
"version": "\u003c 1.3834.0"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "The Claude Desktop app gives you Claude Code with a graphical interface built for running multiple sessions side by side. Prior to 1.3834.0, the CoworkVMService component in Claude Desktop for Windows ran as SYSTEM and did not validate whether the VM bundle directory was a real directory or an NTFS directory junction before creating files within it. A local non-elevated user could replace the user-writable VM bundle directory with a directory junction pointing to an attacker-chosen location, causing the service to create a SYSTEM-owned file in an arbitrary directory. This could be leveraged for local privilege escalation. This vulnerability is fixed in 1.3834.0."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "LOCAL",
"baseScore": 8.5,
"baseSeverity": "HIGH",
"privilegesRequired": "LOW",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:L/AC:L/AT:N/PR:L/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "HIGH",
"vulnIntegrityImpact": "HIGH"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-59",
"description": "CWE-59: Improper Link Resolution Before File Access (\u0027Link Following\u0027)",
"lang": "en",
"type": "CWE"
}
]
},
{
"descriptions": [
{
"cweId": "CWE-269",
"description": "CWE-269: Improper Privilege Management",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-05-13T15:41:48.424Z",
"orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"shortName": "GitHub_M"
},
"references": [
{
"name": "https://github.com/anthropics/claude-code/security/advisories/GHSA-5p5x-5294-qhp3",
"tags": [
"x_refsource_CONFIRM"
],
"url": "https://github.com/anthropics/claude-code/security/advisories/GHSA-5p5x-5294-qhp3"
}
],
"source": {
"advisory": "GHSA-5p5x-5294-qhp3",
"discovery": "UNKNOWN"
},
"title": "Claude Desktop: Local Privilege Escalation via Directory Junction in CoworkVMService"
}
},
"cveMetadata": {
"assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"assignerShortName": "GitHub_M",
"cveId": "CVE-2026-44470",
"datePublished": "2026-05-13T15:41:48.424Z",
"dateReserved": "2026-05-06T17:18:51.782Z",
"dateUpdated": "2026-05-13T18:37:19.139Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-44467 (GCVE-0-2026-44467)
Vulnerability from cvelistv5 – Published: 2026-05-13 15:40 – Updated: 2026-05-14 18:29
VLAI
Title
Claude Desktop: SSH Host Key Verification Bypass Allows Man-in-the-Middle Attack on Remote Sessions
Summary
The Claude Desktop app gives you Claude Code with a graphical interface built for running multiple sessions side by side. From 1.2581.0 to before 1.4304.0, Claude Desktop's SSH remote development feature verified only whether a hostname existed in ~/.ssh/known_hosts without comparing the server's presented host key against the stored key. This allowed a network-positioned attacker to present an arbitrary SSH host key and have the connection silently accepted, enabling a man-in-the-middle attack on remote development sessions. Successful exploitation required the attacker to be in a network position to intercept SSH traffic (e.g., via ARP spoofing, rogue Wi-Fi, or DNS poisoning) and the target hostname to already have an entry in the victim's known_hosts file. This vulnerability is fixed in 1.4304.0.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: total
CISA Coordinator (v2.0.3)
CWE
Assigner
References
1 reference
| URL | Tags |
|---|---|
| https://github.com/anthropics/claude-code/securit… | x_refsource_CONFIRM |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| anthropics | claude-code |
Affected:
>= 1.2581.0, < 1.4304.0
|
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-44467",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "total"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-05-14T18:28:58.122969Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-05-14T18:29:10.605Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"product": "claude-code",
"vendor": "anthropics",
"versions": [
{
"status": "affected",
"version": "\u003e= 1.2581.0, \u003c 1.4304.0"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "The Claude Desktop app gives you Claude Code with a graphical interface built for running multiple sessions side by side. From 1.2581.0 to before 1.4304.0, Claude Desktop\u0027s SSH remote development feature verified only whether a hostname existed in ~/.ssh/known_hosts without comparing the server\u0027s presented host key against the stored key. This allowed a network-positioned attacker to present an arbitrary SSH host key and have the connection silently accepted, enabling a man-in-the-middle attack on remote development sessions. Successful exploitation required the attacker to be in a network position to intercept SSH traffic (e.g., via ARP spoofing, rogue Wi-Fi, or DNS poisoning) and the target hostname to already have an entry in the victim\u0027s known_hosts file. This vulnerability is fixed in 1.4304.0."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "PRESENT",
"attackVector": "ADJACENT",
"baseScore": 7.4,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "PASSIVE",
"vectorString": "CVSS:4.0/AV:A/AC:L/AT:P/PR:N/UI:P/VC:H/VI:H/VA:N/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "NONE",
"vulnConfidentialityImpact": "HIGH",
"vulnIntegrityImpact": "HIGH"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-297",
"description": "CWE-297: Improper Validation of Certificate with Host Mismatch",
"lang": "en",
"type": "CWE"
}
]
},
{
"descriptions": [
{
"cweId": "CWE-322",
"description": "CWE-322: Key Exchange without Entity Authentication",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-05-13T15:40:42.216Z",
"orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"shortName": "GitHub_M"
},
"references": [
{
"name": "https://github.com/anthropics/claude-code/security/advisories/GHSA-3rwf-2g6p-c2f9",
"tags": [
"x_refsource_CONFIRM"
],
"url": "https://github.com/anthropics/claude-code/security/advisories/GHSA-3rwf-2g6p-c2f9"
}
],
"source": {
"advisory": "GHSA-3rwf-2g6p-c2f9",
"discovery": "UNKNOWN"
},
"title": "Claude Desktop: SSH Host Key Verification Bypass Allows Man-in-the-Middle Attack on Remote Sessions"
}
},
"cveMetadata": {
"assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"assignerShortName": "GitHub_M",
"cveId": "CVE-2026-44467",
"datePublished": "2026-05-13T15:40:42.216Z",
"dateReserved": "2026-05-06T15:49:25.193Z",
"dateUpdated": "2026-05-14T18:29:10.605Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-40068 (GCVE-0-2026-40068)
Vulnerability from cvelistv5 – Published: 2026-05-05 20:52 – Updated: 2026-05-06 15:24
VLAI
Title
Claude Code arbitrary code execution via git worktree commondir trust dialog bypass
Summary
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
SSVC
Exploitation: none
Automatable: no
Technical Impact: total
CISA Coordinator (v2.0.3)
CWE
Assigner
References
1 reference
| URL | Tags |
|---|---|
| https://github.com/anthropics/claude-code/securit… | x_refsource_CONFIRM |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| anthropics | claude-code |
Affected:
>= 2.1.63, < 2.1.84
|
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-40068",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "total"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-05-06T14:50:01.382886Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-05-06T15:24:52.454Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"product": "claude-code",
"vendor": "anthropics",
"versions": [
{
"status": "affected",
"version": "\u003e= 2.1.63, \u003c 2.1.84"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "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."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "PRESENT",
"attackVector": "NETWORK",
"baseScore": 7.7,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "PASSIVE",
"vectorString": "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",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "HIGH",
"vulnIntegrityImpact": "HIGH"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-20",
"description": "CWE-20: Improper Input Validation",
"lang": "en",
"type": "CWE"
}
]
},
{
"descriptions": [
{
"cweId": "CWE-77",
"description": "CWE-77: Improper Neutralization of Special Elements used in a Command (\u0027Command Injection\u0027)",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-05-05T20:52:26.089Z",
"orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"shortName": "GitHub_M"
},
"references": [
{
"name": "https://github.com/anthropics/claude-code/security/advisories/GHSA-q5hj-mxqh-vv77",
"tags": [
"x_refsource_CONFIRM"
],
"url": "https://github.com/anthropics/claude-code/security/advisories/GHSA-q5hj-mxqh-vv77"
}
],
"source": {
"advisory": "GHSA-q5hj-mxqh-vv77",
"discovery": "UNKNOWN"
},
"title": "Claude Code arbitrary code execution via git worktree commondir trust dialog bypass"
}
},
"cveMetadata": {
"assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"assignerShortName": "GitHub_M",
"cveId": "CVE-2026-40068",
"datePublished": "2026-05-05T20:52:26.089Z",
"dateReserved": "2026-04-09T00:39:12.204Z",
"dateUpdated": "2026-05-06T15:24:52.454Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-41686 (GCVE-0-2026-41686)
Vulnerability from cvelistv5 – Published: 2026-05-04 18:41 – Updated: 2026-05-05 14:38
VLAI
Title
Claude SDK for TypeScript has Insecure Default File Permissions in Local Filesystem Memory Tool
Summary
Claude SDK for TypeScript provides access to the Claude API from server-side TypeScript or JavaScript applications. From version 0.79.0 to before version 0.91.1, the BetaLocalFilesystemMemoryTool in the Anthropic TypeScript SDK created memory files and directories using the Node.js default modes (0o666 for files, 0o777 for directories), leaving them world-readable on systems with a standard umask and world-writable in environments with a permissive umask such as many Docker base images. A local attacker on a shared host could read persisted agent state, and in containerized deployments could modify memory files to influence subsequent model behavior. This issue has been patched in version 0.91.1.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: partial
CISA Coordinator (v2.0.3)
CWE
- CWE-732 - Incorrect Permission Assignment for Critical Resource
Assigner
References
1 reference
| URL | Tags |
|---|---|
| https://github.com/anthropics/anthropic-sdk-types… | x_refsource_CONFIRM |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| anthropics | anthropic-sdk-typescript |
Affected:
>= 0.79.0, < 0.91.1
|
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-41686",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-05-05T14:35:10.745726Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-05-05T14:38:23.485Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"product": "anthropic-sdk-typescript",
"vendor": "anthropics",
"versions": [
{
"status": "affected",
"version": "\u003e= 0.79.0, \u003c 0.91.1"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "Claude SDK for TypeScript provides access to the Claude API from server-side TypeScript or JavaScript applications. From version 0.79.0 to before version 0.91.1, the BetaLocalFilesystemMemoryTool in the Anthropic TypeScript SDK created memory files and directories using the Node.js default modes (0o666 for files, 0o777 for directories), leaving them world-readable on systems with a standard umask and world-writable in environments with a permissive umask such as many Docker base images. A local attacker on a shared host could read persisted agent state, and in containerized deployments could modify memory files to influence subsequent model behavior. This issue has been patched in version 0.91.1."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "LOCAL",
"baseScore": 4.8,
"baseSeverity": "MEDIUM",
"privilegesRequired": "LOW",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:L/AC:L/AT:N/PR:L/UI:N/VC:L/VI:L/VA:N/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "NONE",
"vulnConfidentialityImpact": "LOW",
"vulnIntegrityImpact": "LOW"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-732",
"description": "CWE-732: Incorrect Permission Assignment for Critical Resource",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-05-04T18:41:40.183Z",
"orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"shortName": "GitHub_M"
},
"references": [
{
"name": "https://github.com/anthropics/anthropic-sdk-typescript/security/advisories/GHSA-p7fg-763f-g4gf",
"tags": [
"x_refsource_CONFIRM"
],
"url": "https://github.com/anthropics/anthropic-sdk-typescript/security/advisories/GHSA-p7fg-763f-g4gf"
}
],
"source": {
"advisory": "GHSA-p7fg-763f-g4gf",
"discovery": "UNKNOWN"
},
"title": "Claude SDK for TypeScript has Insecure Default File Permissions in Local Filesystem Memory Tool"
}
},
"cveMetadata": {
"assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"assignerShortName": "GitHub_M",
"cveId": "CVE-2026-41686",
"datePublished": "2026-05-04T18:41:40.183Z",
"dateReserved": "2026-04-22T03:53:24.406Z",
"dateUpdated": "2026-05-05T14:38:23.485Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-39861 (GCVE-0-2026-39861)
Vulnerability from cvelistv5 – Published: 2026-04-21 00:56 – Updated: 2026-04-21 13:44
VLAI
Title
Claude Code: Sandbox Escape via Symlink Following Allows Arbitrary File Write Outside Workspace
Summary
Claude Code is an agentic coding tool. Prior to version 2.1.64, Claude Code's sandbox did not prevent sandboxed processes from creating symlinks pointing to locations outside the workspace. When Claude Code subsequently wrote to a path within such a symlink, its unsandboxed process followed the symlink and wrote to the target location outside the workspace without prompting the user for confirmation. This allowed a sandbox escape where neither the sandboxed command nor the unsandboxed app could independently write outside the workspace, but their combination could write to arbitrary locations, potentially leading to code execution outside the sandbox. Reliably exploiting this required the ability to add untrusted content into a Claude Code context window to trigger sandboxed code execution via prompt injection. Users on standard Claude Code auto-update have received this fix automatically. Users performing manual updates are advised to update to version 2.1.64 or later.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: total
CISA Coordinator (v2.0.3)
CWE
Assigner
References
1 reference
| URL | Tags |
|---|---|
| https://github.com/anthropics/claude-code/securit… | x_refsource_CONFIRM |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| anthropics | claude-code |
Affected:
< 2.1.64
|
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-39861",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "total"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-04-21T13:44:29.934669Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-04-21T13:44:49.618Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"product": "claude-code",
"vendor": "anthropics",
"versions": [
{
"status": "affected",
"version": "\u003c 2.1.64"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "Claude Code is an agentic coding tool. Prior to version 2.1.64, Claude Code\u0027s sandbox did not prevent sandboxed processes from creating symlinks pointing to locations outside the workspace. When Claude Code subsequently wrote to a path within such a symlink, its unsandboxed process followed the symlink and wrote to the target location outside the workspace without prompting the user for confirmation. This allowed a sandbox escape where neither the sandboxed command nor the unsandboxed app could independently write outside the workspace, but their combination could write to arbitrary locations, potentially leading to code execution outside the sandbox. Reliably exploiting this required the ability to add untrusted content into a Claude Code context window to trigger sandboxed code execution via prompt injection. Users on standard Claude Code auto-update have received this fix automatically. Users performing manual updates are advised to update to version 2.1.64 or later."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "PRESENT",
"attackVector": "NETWORK",
"baseScore": 7.7,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "PASSIVE",
"vectorString": "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",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "HIGH",
"vulnIntegrityImpact": "HIGH"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-22",
"description": "CWE-22: Improper Limitation of a Pathname to a Restricted Directory (\u0027Path Traversal\u0027)",
"lang": "en",
"type": "CWE"
}
]
},
{
"descriptions": [
{
"cweId": "CWE-61",
"description": "CWE-61: UNIX Symbolic Link (Symlink) Following",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-04-21T00:56:39.062Z",
"orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"shortName": "GitHub_M"
},
"references": [
{
"name": "https://github.com/anthropics/claude-code/security/advisories/GHSA-vp62-r36r-9xqp",
"tags": [
"x_refsource_CONFIRM"
],
"url": "https://github.com/anthropics/claude-code/security/advisories/GHSA-vp62-r36r-9xqp"
}
],
"source": {
"advisory": "GHSA-vp62-r36r-9xqp",
"discovery": "UNKNOWN"
},
"title": "Claude Code: Sandbox Escape via Symlink Following Allows Arbitrary File Write Outside Workspace"
}
},
"cveMetadata": {
"assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"assignerShortName": "GitHub_M",
"cveId": "CVE-2026-39861",
"datePublished": "2026-04-21T00:56:39.062Z",
"dateReserved": "2026-04-07T19:13:20.379Z",
"dateUpdated": "2026-04-21T13:44:49.618Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-35603 (GCVE-0-2026-35603)
Vulnerability from cvelistv5 – Published: 2026-04-17 20:38 – Updated: 2026-04-20 14:57
VLAI
Title
Claude Code: Insecure System-Wide Configuration Loading Enables Local Privilege Escalation on Windows
Summary
Claude Code is an agentic coding tool. In versions prior to 2.1.75 on Windows, Claude Code loaded the system-wide default configuration from C:\ProgramData\ClaudeCode\managed-settings.json without validating directory ownership or access permissions. Because the ProgramData directory is writable by non-administrative users by default and the ClaudeCode subdirectory was not pre-created or access-restricted, a low-privileged local user could create this directory and place a malicious configuration file that would be automatically loaded for any user launching Claude Code on the same machine. Exploiting this would have required a shared multi-user Windows system and a victim user to launch Claude Code after the malicious configuration was placed. This issue has been fixed on version 2.1.75.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: total
CISA Coordinator (v2.0.3)
CWE
- CWE-426 - Untrusted Search Path
Assigner
References
1 reference
| URL | Tags |
|---|---|
| https://github.com/anthropics/claude-code/securit… | x_refsource_CONFIRM |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| anthropics | claude-code |
Affected:
< 2.1.75
|
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-35603",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "total"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-04-20T14:52:20.978519Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-04-20T14:57:47.669Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"product": "claude-code",
"vendor": "anthropics",
"versions": [
{
"status": "affected",
"version": "\u003c 2.1.75"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "Claude Code is an agentic coding tool. In versions prior to 2.1.75 on Windows, Claude Code loaded the system-wide default configuration from C:\\ProgramData\\ClaudeCode\\managed-settings.json without validating directory ownership or access permissions. Because the ProgramData directory is writable by non-administrative users by default and the ClaudeCode subdirectory was not pre-created or access-restricted, a low-privileged local user could create this directory and place a malicious configuration file that would be automatically loaded for any user launching Claude Code on the same machine. Exploiting this would have required a shared multi-user Windows system and a victim user to launch Claude Code after the malicious configuration was placed. This issue has been fixed on version 2.1.75."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "PRESENT",
"attackVector": "LOCAL",
"baseScore": 5.4,
"baseSeverity": "MEDIUM",
"privilegesRequired": "LOW",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "PASSIVE",
"vectorString": "CVSS:4.0/AV:L/AC:L/AT:P/PR:L/UI:P/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "HIGH",
"vulnIntegrityImpact": "HIGH"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-426",
"description": "CWE-426: Untrusted Search Path",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-04-17T20:38:49.901Z",
"orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"shortName": "GitHub_M"
},
"references": [
{
"name": "https://github.com/anthropics/claude-code/security/advisories/GHSA-5cwg-9f6j-9jvx",
"tags": [
"x_refsource_CONFIRM"
],
"url": "https://github.com/anthropics/claude-code/security/advisories/GHSA-5cwg-9f6j-9jvx"
}
],
"source": {
"advisory": "GHSA-5cwg-9f6j-9jvx",
"discovery": "UNKNOWN"
},
"title": "Claude Code: Insecure System-Wide Configuration Loading Enables Local Privilege Escalation on Windows"
}
},
"cveMetadata": {
"assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"assignerShortName": "GitHub_M",
"cveId": "CVE-2026-35603",
"datePublished": "2026-04-17T20:38:49.901Z",
"dateReserved": "2026-04-03T21:25:12.162Z",
"dateUpdated": "2026-04-20T14:57:47.669Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-35022 (GCVE-0-2026-35022)
Vulnerability from cvelistv5 – Published: 2026-04-06 18:59 – Updated: 2026-05-29 16:26
VLAI
This CVE ID has been rejected by its CVE Numbering Authority (CNA). It was determined that the -p flag behavior is documented in Anthropic's claude -h output with an explicit warning that non-interactive mode should only be used in trusted directories, making this intended and described behavior rather than a vulnerability.
Show details on NVD website{
"containers": {
"cna": {
"providerMetadata": {
"dateUpdated": "2026-05-29T16:26:48.962Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"rejectedReasons": [
{
"lang": "en",
"supportingMedia": [
{
"base64": false,
"type": "text/html",
"value": "This CVE ID has been rejected by its CVE Numbering Authority (CNA). It was determined that the -p flag behavior is documented in Anthropic\u0027s claude -h output with an explicit warning that non-interactive mode should only be used in trusted directories, making this intended and described behavior rather than a vulnerability."
}
],
"value": "This CVE ID has been rejected by its CVE Numbering Authority (CNA). It was determined that the -p flag behavior is documented in Anthropic\u0027s claude -h output with an explicit warning that non-interactive mode should only be used in trusted directories, making this intended and described behavior rather than a vulnerability."
}
],
"x_generator": {
"engine": "Vulnogram 1.0.2"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-35022",
"datePublished": "2026-04-06T18:59:29.656Z",
"dateRejected": "2026-05-29T16:26:48.962Z",
"dateReserved": "2026-03-31T20:40:15.618Z",
"dateUpdated": "2026-05-29T16:26:48.962Z",
"state": "REJECTED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-35021 (GCVE-0-2026-35021)
Vulnerability from cvelistv5 – Published: 2026-04-06 18:59 – Updated: 2026-05-29 16:26
VLAI
This CVE ID has been rejected by its CVE Numbering Authority (CNA). It was determined that the affected code path cannot be triggered through normal usage of Claude Code.
Show details on NVD website{
"containers": {
"cna": {
"providerMetadata": {
"dateUpdated": "2026-05-29T16:26:11.666Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"rejectedReasons": [
{
"lang": "en",
"supportingMedia": [
{
"base64": false,
"type": "text/html",
"value": "This CVE ID has been rejected by its CVE Numbering Authority (CNA). It was determined that the affected code path cannot be triggered through normal usage of Claude Code."
}
],
"value": "This CVE ID has been rejected by its CVE Numbering Authority (CNA). It was determined that the affected code path cannot be triggered through normal usage of Claude Code."
}
],
"x_generator": {
"engine": "Vulnogram 1.0.2"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-35021",
"datePublished": "2026-04-06T18:59:06.979Z",
"dateRejected": "2026-05-29T16:26:11.666Z",
"dateReserved": "2026-03-31T20:40:15.618Z",
"dateUpdated": "2026-05-29T16:26:11.666Z",
"state": "REJECTED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-35020 (GCVE-0-2026-35020)
Vulnerability from cvelistv5 – Published: 2026-04-06 18:58 – Updated: 2026-05-29 16:25
VLAI
This CVE ID has been rejected by the its CVE Numbering Authority (CNA). It was determined that the attack requires an attacker to already control arbitrary environment variables, a level of access they consider functionally equivalent to code execution and outside the threat model of CLI tools.
Show details on NVD website{
"containers": {
"cna": {
"providerMetadata": {
"dateUpdated": "2026-05-29T16:25:15.369Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"rejectedReasons": [
{
"lang": "en",
"supportingMedia": [
{
"base64": false,
"type": "text/html",
"value": "This CVE ID has been rejected by the its CVE Numbering Authority (CNA). It was determined that the attack requires an attacker to already control arbitrary environment variables, a level of access they consider functionally equivalent to code execution and outside the threat model of CLI tools."
}
],
"value": "This CVE ID has been rejected by the its CVE Numbering Authority (CNA). It was determined that the attack requires an attacker to already control arbitrary environment variables, a level of access they consider functionally equivalent to code execution and outside the threat model of CLI tools."
}
],
"x_generator": {
"engine": "Vulnogram 1.0.2"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-35020",
"datePublished": "2026-04-06T18:58:40.513Z",
"dateRejected": "2026-05-29T16:25:15.369Z",
"dateReserved": "2026-03-31T20:40:15.618Z",
"dateUpdated": "2026-05-29T16:25:15.369Z",
"state": "REJECTED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-34451 (GCVE-0-2026-34451)
Vulnerability from cvelistv5 – Published: 2026-03-31 21:35 – Updated: 2026-04-01 18:57
VLAI
Title
Claude SDK for TypeScript: Memory Tool Path Validation Allows Sandbox Escape to Sibling Directories
Summary
Claude SDK for TypeScript provides access to the Claude API from server-side TypeScript or JavaScript applications. From version 0.79.0 to before version 0.81.0, the local filesystem memory tool in the Anthropic TypeScript SDK validated model-supplied paths using a string prefix check that did not append a trailing path separator. A model steered by prompt injection could supply a crafted path that resolved to a sibling directory sharing the memory root's name as a prefix, allowing reads and writes outside the sandboxed memory directory. This issue has been patched in version 0.81.0.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: partial
CISA Coordinator (v2.0.3)
CWE
Assigner
References
3 references
| URL | Tags |
|---|---|
| https://github.com/anthropics/anthropic-sdk-types… | x_refsource_CONFIRM |
| https://github.com/anthropics/anthropic-sdk-types… | x_refsource_MISC |
| https://github.com/anthropics/anthropic-sdk-types… | x_refsource_MISC |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| anthropics | anthropic-sdk-typescript |
Affected:
>= 0.79.0, < 0.81.0
|
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-34451",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-04-01T18:56:48.666585Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-04-01T18:57:05.442Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"product": "anthropic-sdk-typescript",
"vendor": "anthropics",
"versions": [
{
"status": "affected",
"version": "\u003e= 0.79.0, \u003c 0.81.0"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "Claude SDK for TypeScript provides access to the Claude API from server-side TypeScript or JavaScript applications. From version 0.79.0 to before version 0.81.0, the local filesystem memory tool in the Anthropic TypeScript SDK validated model-supplied paths using a string prefix check that did not append a trailing path separator. A model steered by prompt injection could supply a crafted path that resolved to a sibling directory sharing the memory root\u0027s name as a prefix, allowing reads and writes outside the sandboxed memory directory. This issue has been patched in version 0.81.0."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "HIGH",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 6.3,
"baseSeverity": "MEDIUM",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:H/AT:N/PR:N/UI:N/VC:L/VI:L/VA:N/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "NONE",
"vulnConfidentialityImpact": "LOW",
"vulnIntegrityImpact": "LOW"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-22",
"description": "CWE-22: Improper Limitation of a Pathname to a Restricted Directory (\u0027Path Traversal\u0027)",
"lang": "en",
"type": "CWE"
}
]
},
{
"descriptions": [
{
"cweId": "CWE-41",
"description": "CWE-41: Improper Resolution of Path Equivalence",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-03-31T21:35:21.018Z",
"orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"shortName": "GitHub_M"
},
"references": [
{
"name": "https://github.com/anthropics/anthropic-sdk-typescript/security/advisories/GHSA-5474-4w2j-mq4c",
"tags": [
"x_refsource_CONFIRM"
],
"url": "https://github.com/anthropics/anthropic-sdk-typescript/security/advisories/GHSA-5474-4w2j-mq4c"
},
{
"name": "https://github.com/anthropics/anthropic-sdk-typescript/commit/0ac69b3438ee9c96b21a7d3c39c07b7cdb6995d9",
"tags": [
"x_refsource_MISC"
],
"url": "https://github.com/anthropics/anthropic-sdk-typescript/commit/0ac69b3438ee9c96b21a7d3c39c07b7cdb6995d9"
},
{
"name": "https://github.com/anthropics/anthropic-sdk-typescript/releases/tag/sdk-v0.81.0",
"tags": [
"x_refsource_MISC"
],
"url": "https://github.com/anthropics/anthropic-sdk-typescript/releases/tag/sdk-v0.81.0"
}
],
"source": {
"advisory": "GHSA-5474-4w2j-mq4c",
"discovery": "UNKNOWN"
},
"title": "Claude SDK for TypeScript: Memory Tool Path Validation Allows Sandbox Escape to Sibling Directories"
}
},
"cveMetadata": {
"assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"assignerShortName": "GitHub_M",
"cveId": "CVE-2026-34451",
"datePublished": "2026-03-31T21:35:21.018Z",
"dateReserved": "2026-03-27T18:18:14.895Z",
"dateUpdated": "2026-04-01T18:57:05.442Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-34450 (GCVE-0-2026-34450)
Vulnerability from cvelistv5 – Published: 2026-03-31 21:32 – Updated: 2026-04-01 13:35
VLAI
Title
Claude SDK for Python: Insecure Default File Permissions in Local Filesystem Memory Tool
Summary
The Claude SDK for Python provides access to the Claude API from Python applications. From version 0.86.0 to before version 0.87.0, the local filesystem memory tool in the Anthropic Python SDK created memory files with mode 0o666, leaving them world-readable on systems with a standard umask and world-writable in environments with a permissive umask such as many Docker base images. A local attacker on a shared host could read persisted agent state, and in containerized deployments could modify memory files to influence subsequent model behavior. Both the synchronous and asynchronous memory tool implementations were affected. This issue has been patched in version 0.87.0.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: partial
CISA Coordinator (v2.0.3)
CWE
Assigner
References
3 references
| URL | Tags |
|---|---|
| https://github.com/anthropics/anthropic-sdk-pytho… | x_refsource_CONFIRM |
| https://github.com/anthropics/anthropic-sdk-pytho… | x_refsource_MISC |
| https://github.com/anthropics/anthropic-sdk-pytho… | x_refsource_MISC |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| anthropics | anthropic-sdk-python |
Affected:
>= 0.86.0, < 0.87.0
|
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-34450",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-04-01T13:35:12.188726Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-04-01T13:35:18.879Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"product": "anthropic-sdk-python",
"vendor": "anthropics",
"versions": [
{
"status": "affected",
"version": "\u003e= 0.86.0, \u003c 0.87.0"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "The Claude SDK for Python provides access to the Claude API from Python applications. From version 0.86.0 to before version 0.87.0, the local filesystem memory tool in the Anthropic Python SDK created memory files with mode 0o666, leaving them world-readable on systems with a standard umask and world-writable in environments with a permissive umask such as many Docker base images. A local attacker on a shared host could read persisted agent state, and in containerized deployments could modify memory files to influence subsequent model behavior. Both the synchronous and asynchronous memory tool implementations were affected. This issue has been patched in version 0.87.0."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "LOCAL",
"baseScore": 4.8,
"baseSeverity": "MEDIUM",
"privilegesRequired": "LOW",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:L/AC:L/AT:N/PR:L/UI:N/VC:L/VI:L/VA:N/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "NONE",
"vulnConfidentialityImpact": "LOW",
"vulnIntegrityImpact": "LOW"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-276",
"description": "CWE-276: Incorrect Default Permissions",
"lang": "en",
"type": "CWE"
}
]
},
{
"descriptions": [
{
"cweId": "CWE-732",
"description": "CWE-732: Incorrect Permission Assignment for Critical Resource",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-03-31T21:32:53.617Z",
"orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"shortName": "GitHub_M"
},
"references": [
{
"name": "https://github.com/anthropics/anthropic-sdk-python/security/advisories/GHSA-q5f5-3gjm-7mfm",
"tags": [
"x_refsource_CONFIRM"
],
"url": "https://github.com/anthropics/anthropic-sdk-python/security/advisories/GHSA-q5f5-3gjm-7mfm"
},
{
"name": "https://github.com/anthropics/anthropic-sdk-python/commit/715030ceb4d6dd8d3546e999c680e29532bf1255",
"tags": [
"x_refsource_MISC"
],
"url": "https://github.com/anthropics/anthropic-sdk-python/commit/715030ceb4d6dd8d3546e999c680e29532bf1255"
},
{
"name": "https://github.com/anthropics/anthropic-sdk-python/releases/tag/v0.87.0",
"tags": [
"x_refsource_MISC"
],
"url": "https://github.com/anthropics/anthropic-sdk-python/releases/tag/v0.87.0"
}
],
"source": {
"advisory": "GHSA-q5f5-3gjm-7mfm",
"discovery": "UNKNOWN"
},
"title": "Claude SDK for Python: Insecure Default File Permissions in Local Filesystem Memory Tool"
}
},
"cveMetadata": {
"assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"assignerShortName": "GitHub_M",
"cveId": "CVE-2026-34450",
"datePublished": "2026-03-31T21:32:53.617Z",
"dateReserved": "2026-03-27T18:18:14.895Z",
"dateUpdated": "2026-04-01T13:35:18.879Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-34452 (GCVE-0-2026-34452)
Vulnerability from cvelistv5 – Published: 2026-03-31 21:32 – Updated: 2026-04-03 16:08
VLAI
Title
Claude SDK for Python: Memory Tool Path Validation Race Condition Allows Sandbox Escape
Summary
The Claude SDK for Python provides access to the Claude API from Python applications. From version 0.86.0 to before version 0.87.0, the async local filesystem memory tool in the Anthropic Python SDK validated that model-supplied paths resolved inside the sandboxed memory directory, but then returned the unresolved path for subsequent file operations. A local attacker able to write to the memory directory could retarget a symlink between validation and use, causing reads or writes to escape the sandbox. The synchronous memory tool implementation was not affected. This issue has been patched in version 0.87.0.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: partial
CISA Coordinator (v2.0.3)
CWE
Assigner
References
3 references
| URL | Tags |
|---|---|
| https://github.com/anthropics/anthropic-sdk-pytho… | x_refsource_CONFIRM |
| https://github.com/anthropics/anthropic-sdk-pytho… | x_refsource_MISC |
| https://github.com/anthropics/anthropic-sdk-pytho… | x_refsource_MISC |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| anthropics | anthropic-sdk-python |
Affected:
>= 0.86.0, < 0.87.0
|
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-34452",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-04-03T16:06:47.039006Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-04-03T16:08:25.309Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"product": "anthropic-sdk-python",
"vendor": "anthropics",
"versions": [
{
"status": "affected",
"version": "\u003e= 0.86.0, \u003c 0.87.0"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "The Claude SDK for Python provides access to the Claude API from Python applications. From version 0.86.0 to before version 0.87.0, the async local filesystem memory tool in the Anthropic Python SDK validated that model-supplied paths resolved inside the sandboxed memory directory, but then returned the unresolved path for subsequent file operations. A local attacker able to write to the memory directory could retarget a symlink between validation and use, causing reads or writes to escape the sandbox. The synchronous memory tool implementation was not affected. This issue has been patched in version 0.87.0."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "HIGH",
"attackRequirements": "NONE",
"attackVector": "LOCAL",
"baseScore": 5.8,
"baseSeverity": "MEDIUM",
"privilegesRequired": "LOW",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:L/AC:H/AT:N/PR:L/UI:N/VC:H/VI:L/VA:N/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "NONE",
"vulnConfidentialityImpact": "HIGH",
"vulnIntegrityImpact": "LOW"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-59",
"description": "CWE-59: Improper Link Resolution Before File Access (\u0027Link Following\u0027)",
"lang": "en",
"type": "CWE"
}
]
},
{
"descriptions": [
{
"cweId": "CWE-367",
"description": "CWE-367: Time-of-check Time-of-use (TOCTOU) Race Condition",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-03-31T21:32:37.537Z",
"orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"shortName": "GitHub_M"
},
"references": [
{
"name": "https://github.com/anthropics/anthropic-sdk-python/security/advisories/GHSA-w828-4qhx-vxx3",
"tags": [
"x_refsource_CONFIRM"
],
"url": "https://github.com/anthropics/anthropic-sdk-python/security/advisories/GHSA-w828-4qhx-vxx3"
},
{
"name": "https://github.com/anthropics/anthropic-sdk-python/commit/6599043eee6e86dce16953fcd1fd828052052be6",
"tags": [
"x_refsource_MISC"
],
"url": "https://github.com/anthropics/anthropic-sdk-python/commit/6599043eee6e86dce16953fcd1fd828052052be6"
},
{
"name": "https://github.com/anthropics/anthropic-sdk-python/releases/tag/v0.87.0",
"tags": [
"x_refsource_MISC"
],
"url": "https://github.com/anthropics/anthropic-sdk-python/releases/tag/v0.87.0"
}
],
"source": {
"advisory": "GHSA-w828-4qhx-vxx3",
"discovery": "UNKNOWN"
},
"title": "Claude SDK for Python: Memory Tool Path Validation Race Condition Allows Sandbox Escape"
}
},
"cveMetadata": {
"assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"assignerShortName": "GitHub_M",
"cveId": "CVE-2026-34452",
"datePublished": "2026-03-31T21:32:37.537Z",
"dateReserved": "2026-03-27T18:18:14.895Z",
"dateUpdated": "2026-04-03T16:08:25.309Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-22561 (GCVE-0-2026-22561)
Vulnerability from cvelistv5 – Published: 2026-03-31 15:30 – Updated: 2026-05-10 13:58
VLAI
Summary
Uncontrolled search path elements in Anthropic Claude for Windows installer (Claude Setup.exe) versions prior to 1.1.3363 allow local privilege escalation via DLL search-order hijacking. The installer loads DLLs (e.g., profapi.dll) from its own directory after UAC elevation, enabling arbitrary code execution if a malicious DLL is planted alongside the installer.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: partial
CISA Coordinator (v2.0.3)
CWE
- CWE-427 - Uncontrolled Search Path Element
Assigner
References
1 reference
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| Anthropic | Claude Desktop - Windows |
Affected:
0 , < 1.1.3363
(semver)
|
Credits
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-22561",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-03-31T16:27:37.909291Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-427",
"description": "CWE-427 Uncontrolled Search Path Element",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-05-10T13:58:31.956Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"product": "Claude Desktop - Windows",
"vendor": "Anthropic",
"versions": [
{
"lessThan": "1.1.3363",
"status": "affected",
"version": "0",
"versionType": "semver"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "finder",
"value": "Kazuma Matsumoto, a security reseaercher at GMO Cybersecurity by IERAE, Inc"
}
],
"descriptions": [
{
"lang": "en",
"value": "Uncontrolled search path elements in Anthropic Claude for Windows installer (Claude Setup.exe) versions prior to 1.1.3363 allow local privilege escalation via DLL search-order hijacking. The installer loads DLLs (e.g., profapi.dll) from its own directory after UAC elevation, enabling arbitrary code execution if a malicious DLL is planted alongside the installer."
}
],
"metrics": [
{
"cvssV4_0": {
"baseScore": 4.7,
"baseSeverity": "MEDIUM",
"vectorString": "CVSS:4.0/AV:L/AC:L/AT:N/PR:L/UI:P/VC:N/VI:N/VA:N/SC:H/SI:H/SA:H",
"version": "4.0"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-03-31T18:13:52.328Z",
"orgId": "36234546-b8fa-4601-9d6f-f4e334aa8ea1",
"shortName": "hackerone"
},
"references": [
{
"url": "https://trust.anthropic.com/resources?s=1cvig6ldp3zvuj1yffzr11\u0026name=cve-2026-22561-dll-search-order-hijacking-in-claude-for-windows-installer"
}
]
}
},
"cveMetadata": {
"assignerOrgId": "36234546-b8fa-4601-9d6f-f4e334aa8ea1",
"assignerShortName": "hackerone",
"cveId": "CVE-2026-22561",
"datePublished": "2026-03-31T15:30:38.876Z",
"dateReserved": "2026-01-07T15:39:03.440Z",
"dateUpdated": "2026-05-10T13:58:31.956Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-33068 (GCVE-0-2026-33068)
Vulnerability from cvelistv5 – Published: 2026-03-20 08:17 – Updated: 2026-03-20 13:48
VLAI
Title
Claude Code has a Workspace Trust Dialog Bypass via Repo-Controlled Settings File
Summary
Claude Code is an agentic coding tool. Versions prior to 2.1.53 resolved the permission mode from settings files, including the repo-controlled .claude/settings.json, before determining whether to display the workspace trust confirmation dialog. A malicious repository could set permissions.defaultMode to bypassPermissions in its committed .claude/settings.json, causing the trust dialog to be silently skipped on first open. This allowed a user to be placed into a permissive mode without seeing the trust confirmation prompt, making it easier for an attacker-controlled repository to gain tool execution without explicit user consent. This issue has been patched in version 2.1.53.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: total
CISA Coordinator (v2.0.3)
CWE
- CWE-807 - Reliance on Untrusted Inputs in a Security Decision
Assigner
References
1 reference
| URL | Tags |
|---|---|
| https://github.com/anthropics/claude-code/securit… | x_refsource_CONFIRM |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| anthropics | claude-code |
Affected:
< 2.1.53
|
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-33068",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "total"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-03-20T13:48:28.263574Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-03-20T13:48:36.014Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"product": "claude-code",
"vendor": "anthropics",
"versions": [
{
"status": "affected",
"version": "\u003c 2.1.53"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "Claude Code is an agentic coding tool. Versions prior to 2.1.53 resolved the permission mode from settings files, including the repo-controlled .claude/settings.json, before determining whether to display the workspace trust confirmation dialog. A malicious repository could set permissions.defaultMode to bypassPermissions in its committed .claude/settings.json, causing the trust dialog to be silently skipped on first open. This allowed a user to be placed into a permissive mode without seeing the trust confirmation prompt, making it easier for an attacker-controlled repository to gain tool execution without explicit user consent. This issue has been patched in version 2.1.53."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "PRESENT",
"attackVector": "NETWORK",
"baseScore": 7.7,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "PASSIVE",
"vectorString": "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",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "HIGH",
"vulnIntegrityImpact": "HIGH"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-807",
"description": "CWE-807: Reliance on Untrusted Inputs in a Security Decision",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-03-20T08:17:47.794Z",
"orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"shortName": "GitHub_M"
},
"references": [
{
"name": "https://github.com/anthropics/claude-code/security/advisories/GHSA-mmgp-wc2j-qcv7",
"tags": [
"x_refsource_CONFIRM"
],
"url": "https://github.com/anthropics/claude-code/security/advisories/GHSA-mmgp-wc2j-qcv7"
}
],
"source": {
"advisory": "GHSA-mmgp-wc2j-qcv7",
"discovery": "UNKNOWN"
},
"title": "Claude Code has a Workspace Trust Dialog Bypass via Repo-Controlled Settings File"
}
},
"cveMetadata": {
"assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"assignerShortName": "GitHub_M",
"cveId": "CVE-2026-33068",
"datePublished": "2026-03-20T08:17:47.794Z",
"dateReserved": "2026-03-17T19:27:06.343Z",
"dateUpdated": "2026-03-20T13:48:36.014Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-44470 (GCVE-0-2026-44470)
Vulnerability from nvd – Published: 2026-05-13 15:41 – Updated: 2026-05-13 18:37
VLAI
Title
Claude Desktop: Local Privilege Escalation via Directory Junction in CoworkVMService
Summary
The Claude Desktop app gives you Claude Code with a graphical interface built for running multiple sessions side by side. Prior to 1.3834.0, the CoworkVMService component in Claude Desktop for Windows ran as SYSTEM and did not validate whether the VM bundle directory was a real directory or an NTFS directory junction before creating files within it. A local non-elevated user could replace the user-writable VM bundle directory with a directory junction pointing to an attacker-chosen location, causing the service to create a SYSTEM-owned file in an arbitrary directory. This could be leveraged for local privilege escalation. This vulnerability is fixed in 1.3834.0.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: total
CISA Coordinator (v2.0.3)
CWE
Assigner
References
1 reference
| URL | Tags |
|---|---|
| https://github.com/anthropics/claude-code/securit… | x_refsource_CONFIRM |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| anthropics | claude-code |
Affected:
< 1.3834.0
|
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-44470",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "total"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-05-13T18:36:56.249906Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-05-13T18:37:19.139Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"product": "claude-code",
"vendor": "anthropics",
"versions": [
{
"status": "affected",
"version": "\u003c 1.3834.0"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "The Claude Desktop app gives you Claude Code with a graphical interface built for running multiple sessions side by side. Prior to 1.3834.0, the CoworkVMService component in Claude Desktop for Windows ran as SYSTEM and did not validate whether the VM bundle directory was a real directory or an NTFS directory junction before creating files within it. A local non-elevated user could replace the user-writable VM bundle directory with a directory junction pointing to an attacker-chosen location, causing the service to create a SYSTEM-owned file in an arbitrary directory. This could be leveraged for local privilege escalation. This vulnerability is fixed in 1.3834.0."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "LOCAL",
"baseScore": 8.5,
"baseSeverity": "HIGH",
"privilegesRequired": "LOW",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:L/AC:L/AT:N/PR:L/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "HIGH",
"vulnIntegrityImpact": "HIGH"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-59",
"description": "CWE-59: Improper Link Resolution Before File Access (\u0027Link Following\u0027)",
"lang": "en",
"type": "CWE"
}
]
},
{
"descriptions": [
{
"cweId": "CWE-269",
"description": "CWE-269: Improper Privilege Management",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-05-13T15:41:48.424Z",
"orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"shortName": "GitHub_M"
},
"references": [
{
"name": "https://github.com/anthropics/claude-code/security/advisories/GHSA-5p5x-5294-qhp3",
"tags": [
"x_refsource_CONFIRM"
],
"url": "https://github.com/anthropics/claude-code/security/advisories/GHSA-5p5x-5294-qhp3"
}
],
"source": {
"advisory": "GHSA-5p5x-5294-qhp3",
"discovery": "UNKNOWN"
},
"title": "Claude Desktop: Local Privilege Escalation via Directory Junction in CoworkVMService"
}
},
"cveMetadata": {
"assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"assignerShortName": "GitHub_M",
"cveId": "CVE-2026-44470",
"datePublished": "2026-05-13T15:41:48.424Z",
"dateReserved": "2026-05-06T17:18:51.782Z",
"dateUpdated": "2026-05-13T18:37:19.139Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-44467 (GCVE-0-2026-44467)
Vulnerability from nvd – Published: 2026-05-13 15:40 – Updated: 2026-05-14 18:29
VLAI
Title
Claude Desktop: SSH Host Key Verification Bypass Allows Man-in-the-Middle Attack on Remote Sessions
Summary
The Claude Desktop app gives you Claude Code with a graphical interface built for running multiple sessions side by side. From 1.2581.0 to before 1.4304.0, Claude Desktop's SSH remote development feature verified only whether a hostname existed in ~/.ssh/known_hosts without comparing the server's presented host key against the stored key. This allowed a network-positioned attacker to present an arbitrary SSH host key and have the connection silently accepted, enabling a man-in-the-middle attack on remote development sessions. Successful exploitation required the attacker to be in a network position to intercept SSH traffic (e.g., via ARP spoofing, rogue Wi-Fi, or DNS poisoning) and the target hostname to already have an entry in the victim's known_hosts file. This vulnerability is fixed in 1.4304.0.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: total
CISA Coordinator (v2.0.3)
CWE
Assigner
References
1 reference
| URL | Tags |
|---|---|
| https://github.com/anthropics/claude-code/securit… | x_refsource_CONFIRM |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| anthropics | claude-code |
Affected:
>= 1.2581.0, < 1.4304.0
|
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-44467",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "total"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-05-14T18:28:58.122969Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-05-14T18:29:10.605Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"product": "claude-code",
"vendor": "anthropics",
"versions": [
{
"status": "affected",
"version": "\u003e= 1.2581.0, \u003c 1.4304.0"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "The Claude Desktop app gives you Claude Code with a graphical interface built for running multiple sessions side by side. From 1.2581.0 to before 1.4304.0, Claude Desktop\u0027s SSH remote development feature verified only whether a hostname existed in ~/.ssh/known_hosts without comparing the server\u0027s presented host key against the stored key. This allowed a network-positioned attacker to present an arbitrary SSH host key and have the connection silently accepted, enabling a man-in-the-middle attack on remote development sessions. Successful exploitation required the attacker to be in a network position to intercept SSH traffic (e.g., via ARP spoofing, rogue Wi-Fi, or DNS poisoning) and the target hostname to already have an entry in the victim\u0027s known_hosts file. This vulnerability is fixed in 1.4304.0."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "PRESENT",
"attackVector": "ADJACENT",
"baseScore": 7.4,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "PASSIVE",
"vectorString": "CVSS:4.0/AV:A/AC:L/AT:P/PR:N/UI:P/VC:H/VI:H/VA:N/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "NONE",
"vulnConfidentialityImpact": "HIGH",
"vulnIntegrityImpact": "HIGH"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-297",
"description": "CWE-297: Improper Validation of Certificate with Host Mismatch",
"lang": "en",
"type": "CWE"
}
]
},
{
"descriptions": [
{
"cweId": "CWE-322",
"description": "CWE-322: Key Exchange without Entity Authentication",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-05-13T15:40:42.216Z",
"orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"shortName": "GitHub_M"
},
"references": [
{
"name": "https://github.com/anthropics/claude-code/security/advisories/GHSA-3rwf-2g6p-c2f9",
"tags": [
"x_refsource_CONFIRM"
],
"url": "https://github.com/anthropics/claude-code/security/advisories/GHSA-3rwf-2g6p-c2f9"
}
],
"source": {
"advisory": "GHSA-3rwf-2g6p-c2f9",
"discovery": "UNKNOWN"
},
"title": "Claude Desktop: SSH Host Key Verification Bypass Allows Man-in-the-Middle Attack on Remote Sessions"
}
},
"cveMetadata": {
"assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"assignerShortName": "GitHub_M",
"cveId": "CVE-2026-44467",
"datePublished": "2026-05-13T15:40:42.216Z",
"dateReserved": "2026-05-06T15:49:25.193Z",
"dateUpdated": "2026-05-14T18:29:10.605Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-40068 (GCVE-0-2026-40068)
Vulnerability from nvd – Published: 2026-05-05 20:52 – Updated: 2026-05-06 15:24
VLAI
Title
Claude Code arbitrary code execution via git worktree commondir trust dialog bypass
Summary
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
SSVC
Exploitation: none
Automatable: no
Technical Impact: total
CISA Coordinator (v2.0.3)
CWE
Assigner
References
1 reference
| URL | Tags |
|---|---|
| https://github.com/anthropics/claude-code/securit… | x_refsource_CONFIRM |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| anthropics | claude-code |
Affected:
>= 2.1.63, < 2.1.84
|
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-40068",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "total"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-05-06T14:50:01.382886Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-05-06T15:24:52.454Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"product": "claude-code",
"vendor": "anthropics",
"versions": [
{
"status": "affected",
"version": "\u003e= 2.1.63, \u003c 2.1.84"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "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."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "PRESENT",
"attackVector": "NETWORK",
"baseScore": 7.7,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "PASSIVE",
"vectorString": "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",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "HIGH",
"vulnIntegrityImpact": "HIGH"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-20",
"description": "CWE-20: Improper Input Validation",
"lang": "en",
"type": "CWE"
}
]
},
{
"descriptions": [
{
"cweId": "CWE-77",
"description": "CWE-77: Improper Neutralization of Special Elements used in a Command (\u0027Command Injection\u0027)",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-05-05T20:52:26.089Z",
"orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"shortName": "GitHub_M"
},
"references": [
{
"name": "https://github.com/anthropics/claude-code/security/advisories/GHSA-q5hj-mxqh-vv77",
"tags": [
"x_refsource_CONFIRM"
],
"url": "https://github.com/anthropics/claude-code/security/advisories/GHSA-q5hj-mxqh-vv77"
}
],
"source": {
"advisory": "GHSA-q5hj-mxqh-vv77",
"discovery": "UNKNOWN"
},
"title": "Claude Code arbitrary code execution via git worktree commondir trust dialog bypass"
}
},
"cveMetadata": {
"assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"assignerShortName": "GitHub_M",
"cveId": "CVE-2026-40068",
"datePublished": "2026-05-05T20:52:26.089Z",
"dateReserved": "2026-04-09T00:39:12.204Z",
"dateUpdated": "2026-05-06T15:24:52.454Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-41686 (GCVE-0-2026-41686)
Vulnerability from nvd – Published: 2026-05-04 18:41 – Updated: 2026-05-05 14:38
VLAI
Title
Claude SDK for TypeScript has Insecure Default File Permissions in Local Filesystem Memory Tool
Summary
Claude SDK for TypeScript provides access to the Claude API from server-side TypeScript or JavaScript applications. From version 0.79.0 to before version 0.91.1, the BetaLocalFilesystemMemoryTool in the Anthropic TypeScript SDK created memory files and directories using the Node.js default modes (0o666 for files, 0o777 for directories), leaving them world-readable on systems with a standard umask and world-writable in environments with a permissive umask such as many Docker base images. A local attacker on a shared host could read persisted agent state, and in containerized deployments could modify memory files to influence subsequent model behavior. This issue has been patched in version 0.91.1.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: partial
CISA Coordinator (v2.0.3)
CWE
- CWE-732 - Incorrect Permission Assignment for Critical Resource
Assigner
References
1 reference
| URL | Tags |
|---|---|
| https://github.com/anthropics/anthropic-sdk-types… | x_refsource_CONFIRM |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| anthropics | anthropic-sdk-typescript |
Affected:
>= 0.79.0, < 0.91.1
|
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-41686",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-05-05T14:35:10.745726Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-05-05T14:38:23.485Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"product": "anthropic-sdk-typescript",
"vendor": "anthropics",
"versions": [
{
"status": "affected",
"version": "\u003e= 0.79.0, \u003c 0.91.1"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "Claude SDK for TypeScript provides access to the Claude API from server-side TypeScript or JavaScript applications. From version 0.79.0 to before version 0.91.1, the BetaLocalFilesystemMemoryTool in the Anthropic TypeScript SDK created memory files and directories using the Node.js default modes (0o666 for files, 0o777 for directories), leaving them world-readable on systems with a standard umask and world-writable in environments with a permissive umask such as many Docker base images. A local attacker on a shared host could read persisted agent state, and in containerized deployments could modify memory files to influence subsequent model behavior. This issue has been patched in version 0.91.1."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "LOCAL",
"baseScore": 4.8,
"baseSeverity": "MEDIUM",
"privilegesRequired": "LOW",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:L/AC:L/AT:N/PR:L/UI:N/VC:L/VI:L/VA:N/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "NONE",
"vulnConfidentialityImpact": "LOW",
"vulnIntegrityImpact": "LOW"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-732",
"description": "CWE-732: Incorrect Permission Assignment for Critical Resource",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-05-04T18:41:40.183Z",
"orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"shortName": "GitHub_M"
},
"references": [
{
"name": "https://github.com/anthropics/anthropic-sdk-typescript/security/advisories/GHSA-p7fg-763f-g4gf",
"tags": [
"x_refsource_CONFIRM"
],
"url": "https://github.com/anthropics/anthropic-sdk-typescript/security/advisories/GHSA-p7fg-763f-g4gf"
}
],
"source": {
"advisory": "GHSA-p7fg-763f-g4gf",
"discovery": "UNKNOWN"
},
"title": "Claude SDK for TypeScript has Insecure Default File Permissions in Local Filesystem Memory Tool"
}
},
"cveMetadata": {
"assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"assignerShortName": "GitHub_M",
"cveId": "CVE-2026-41686",
"datePublished": "2026-05-04T18:41:40.183Z",
"dateReserved": "2026-04-22T03:53:24.406Z",
"dateUpdated": "2026-05-05T14:38:23.485Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-39861 (GCVE-0-2026-39861)
Vulnerability from nvd – Published: 2026-04-21 00:56 – Updated: 2026-04-21 13:44
VLAI
Title
Claude Code: Sandbox Escape via Symlink Following Allows Arbitrary File Write Outside Workspace
Summary
Claude Code is an agentic coding tool. Prior to version 2.1.64, Claude Code's sandbox did not prevent sandboxed processes from creating symlinks pointing to locations outside the workspace. When Claude Code subsequently wrote to a path within such a symlink, its unsandboxed process followed the symlink and wrote to the target location outside the workspace without prompting the user for confirmation. This allowed a sandbox escape where neither the sandboxed command nor the unsandboxed app could independently write outside the workspace, but their combination could write to arbitrary locations, potentially leading to code execution outside the sandbox. Reliably exploiting this required the ability to add untrusted content into a Claude Code context window to trigger sandboxed code execution via prompt injection. Users on standard Claude Code auto-update have received this fix automatically. Users performing manual updates are advised to update to version 2.1.64 or later.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: total
CISA Coordinator (v2.0.3)
CWE
Assigner
References
1 reference
| URL | Tags |
|---|---|
| https://github.com/anthropics/claude-code/securit… | x_refsource_CONFIRM |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| anthropics | claude-code |
Affected:
< 2.1.64
|
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-39861",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "total"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-04-21T13:44:29.934669Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-04-21T13:44:49.618Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"product": "claude-code",
"vendor": "anthropics",
"versions": [
{
"status": "affected",
"version": "\u003c 2.1.64"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "Claude Code is an agentic coding tool. Prior to version 2.1.64, Claude Code\u0027s sandbox did not prevent sandboxed processes from creating symlinks pointing to locations outside the workspace. When Claude Code subsequently wrote to a path within such a symlink, its unsandboxed process followed the symlink and wrote to the target location outside the workspace without prompting the user for confirmation. This allowed a sandbox escape where neither the sandboxed command nor the unsandboxed app could independently write outside the workspace, but their combination could write to arbitrary locations, potentially leading to code execution outside the sandbox. Reliably exploiting this required the ability to add untrusted content into a Claude Code context window to trigger sandboxed code execution via prompt injection. Users on standard Claude Code auto-update have received this fix automatically. Users performing manual updates are advised to update to version 2.1.64 or later."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "PRESENT",
"attackVector": "NETWORK",
"baseScore": 7.7,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "PASSIVE",
"vectorString": "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",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "HIGH",
"vulnIntegrityImpact": "HIGH"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-22",
"description": "CWE-22: Improper Limitation of a Pathname to a Restricted Directory (\u0027Path Traversal\u0027)",
"lang": "en",
"type": "CWE"
}
]
},
{
"descriptions": [
{
"cweId": "CWE-61",
"description": "CWE-61: UNIX Symbolic Link (Symlink) Following",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-04-21T00:56:39.062Z",
"orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"shortName": "GitHub_M"
},
"references": [
{
"name": "https://github.com/anthropics/claude-code/security/advisories/GHSA-vp62-r36r-9xqp",
"tags": [
"x_refsource_CONFIRM"
],
"url": "https://github.com/anthropics/claude-code/security/advisories/GHSA-vp62-r36r-9xqp"
}
],
"source": {
"advisory": "GHSA-vp62-r36r-9xqp",
"discovery": "UNKNOWN"
},
"title": "Claude Code: Sandbox Escape via Symlink Following Allows Arbitrary File Write Outside Workspace"
}
},
"cveMetadata": {
"assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"assignerShortName": "GitHub_M",
"cveId": "CVE-2026-39861",
"datePublished": "2026-04-21T00:56:39.062Z",
"dateReserved": "2026-04-07T19:13:20.379Z",
"dateUpdated": "2026-04-21T13:44:49.618Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-35603 (GCVE-0-2026-35603)
Vulnerability from nvd – Published: 2026-04-17 20:38 – Updated: 2026-04-20 14:57
VLAI
Title
Claude Code: Insecure System-Wide Configuration Loading Enables Local Privilege Escalation on Windows
Summary
Claude Code is an agentic coding tool. In versions prior to 2.1.75 on Windows, Claude Code loaded the system-wide default configuration from C:\ProgramData\ClaudeCode\managed-settings.json without validating directory ownership or access permissions. Because the ProgramData directory is writable by non-administrative users by default and the ClaudeCode subdirectory was not pre-created or access-restricted, a low-privileged local user could create this directory and place a malicious configuration file that would be automatically loaded for any user launching Claude Code on the same machine. Exploiting this would have required a shared multi-user Windows system and a victim user to launch Claude Code after the malicious configuration was placed. This issue has been fixed on version 2.1.75.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: total
CISA Coordinator (v2.0.3)
CWE
- CWE-426 - Untrusted Search Path
Assigner
References
1 reference
| URL | Tags |
|---|---|
| https://github.com/anthropics/claude-code/securit… | x_refsource_CONFIRM |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| anthropics | claude-code |
Affected:
< 2.1.75
|
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-35603",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "total"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-04-20T14:52:20.978519Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-04-20T14:57:47.669Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"product": "claude-code",
"vendor": "anthropics",
"versions": [
{
"status": "affected",
"version": "\u003c 2.1.75"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "Claude Code is an agentic coding tool. In versions prior to 2.1.75 on Windows, Claude Code loaded the system-wide default configuration from C:\\ProgramData\\ClaudeCode\\managed-settings.json without validating directory ownership or access permissions. Because the ProgramData directory is writable by non-administrative users by default and the ClaudeCode subdirectory was not pre-created or access-restricted, a low-privileged local user could create this directory and place a malicious configuration file that would be automatically loaded for any user launching Claude Code on the same machine. Exploiting this would have required a shared multi-user Windows system and a victim user to launch Claude Code after the malicious configuration was placed. This issue has been fixed on version 2.1.75."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "PRESENT",
"attackVector": "LOCAL",
"baseScore": 5.4,
"baseSeverity": "MEDIUM",
"privilegesRequired": "LOW",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "PASSIVE",
"vectorString": "CVSS:4.0/AV:L/AC:L/AT:P/PR:L/UI:P/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "HIGH",
"vulnIntegrityImpact": "HIGH"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-426",
"description": "CWE-426: Untrusted Search Path",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-04-17T20:38:49.901Z",
"orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"shortName": "GitHub_M"
},
"references": [
{
"name": "https://github.com/anthropics/claude-code/security/advisories/GHSA-5cwg-9f6j-9jvx",
"tags": [
"x_refsource_CONFIRM"
],
"url": "https://github.com/anthropics/claude-code/security/advisories/GHSA-5cwg-9f6j-9jvx"
}
],
"source": {
"advisory": "GHSA-5cwg-9f6j-9jvx",
"discovery": "UNKNOWN"
},
"title": "Claude Code: Insecure System-Wide Configuration Loading Enables Local Privilege Escalation on Windows"
}
},
"cveMetadata": {
"assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"assignerShortName": "GitHub_M",
"cveId": "CVE-2026-35603",
"datePublished": "2026-04-17T20:38:49.901Z",
"dateReserved": "2026-04-03T21:25:12.162Z",
"dateUpdated": "2026-04-20T14:57:47.669Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-35022 (GCVE-0-2026-35022)
Vulnerability from nvd – Published: 2026-04-06 18:59 – Updated: 2026-05-29 16:26
VLAI
This CVE ID has been rejected by its CVE Numbering Authority (CNA). It was determined that the -p flag behavior is documented in Anthropic's claude -h output with an explicit warning that non-interactive mode should only be used in trusted directories, making this intended and described behavior rather than a vulnerability.
Show details on NVD website{
"containers": {
"cna": {
"providerMetadata": {
"dateUpdated": "2026-05-29T16:26:48.962Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"rejectedReasons": [
{
"lang": "en",
"supportingMedia": [
{
"base64": false,
"type": "text/html",
"value": "This CVE ID has been rejected by its CVE Numbering Authority (CNA). It was determined that the -p flag behavior is documented in Anthropic\u0027s claude -h output with an explicit warning that non-interactive mode should only be used in trusted directories, making this intended and described behavior rather than a vulnerability."
}
],
"value": "This CVE ID has been rejected by its CVE Numbering Authority (CNA). It was determined that the -p flag behavior is documented in Anthropic\u0027s claude -h output with an explicit warning that non-interactive mode should only be used in trusted directories, making this intended and described behavior rather than a vulnerability."
}
],
"x_generator": {
"engine": "Vulnogram 1.0.2"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-35022",
"datePublished": "2026-04-06T18:59:29.656Z",
"dateRejected": "2026-05-29T16:26:48.962Z",
"dateReserved": "2026-03-31T20:40:15.618Z",
"dateUpdated": "2026-05-29T16:26:48.962Z",
"state": "REJECTED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-35021 (GCVE-0-2026-35021)
Vulnerability from nvd – Published: 2026-04-06 18:59 – Updated: 2026-05-29 16:26
VLAI
This CVE ID has been rejected by its CVE Numbering Authority (CNA). It was determined that the affected code path cannot be triggered through normal usage of Claude Code.
Show details on NVD website{
"containers": {
"cna": {
"providerMetadata": {
"dateUpdated": "2026-05-29T16:26:11.666Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"rejectedReasons": [
{
"lang": "en",
"supportingMedia": [
{
"base64": false,
"type": "text/html",
"value": "This CVE ID has been rejected by its CVE Numbering Authority (CNA). It was determined that the affected code path cannot be triggered through normal usage of Claude Code."
}
],
"value": "This CVE ID has been rejected by its CVE Numbering Authority (CNA). It was determined that the affected code path cannot be triggered through normal usage of Claude Code."
}
],
"x_generator": {
"engine": "Vulnogram 1.0.2"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-35021",
"datePublished": "2026-04-06T18:59:06.979Z",
"dateRejected": "2026-05-29T16:26:11.666Z",
"dateReserved": "2026-03-31T20:40:15.618Z",
"dateUpdated": "2026-05-29T16:26:11.666Z",
"state": "REJECTED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-35020 (GCVE-0-2026-35020)
Vulnerability from nvd – Published: 2026-04-06 18:58 – Updated: 2026-05-29 16:25
VLAI
This CVE ID has been rejected by the its CVE Numbering Authority (CNA). It was determined that the attack requires an attacker to already control arbitrary environment variables, a level of access they consider functionally equivalent to code execution and outside the threat model of CLI tools.
Show details on NVD website{
"containers": {
"cna": {
"providerMetadata": {
"dateUpdated": "2026-05-29T16:25:15.369Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"rejectedReasons": [
{
"lang": "en",
"supportingMedia": [
{
"base64": false,
"type": "text/html",
"value": "This CVE ID has been rejected by the its CVE Numbering Authority (CNA). It was determined that the attack requires an attacker to already control arbitrary environment variables, a level of access they consider functionally equivalent to code execution and outside the threat model of CLI tools."
}
],
"value": "This CVE ID has been rejected by the its CVE Numbering Authority (CNA). It was determined that the attack requires an attacker to already control arbitrary environment variables, a level of access they consider functionally equivalent to code execution and outside the threat model of CLI tools."
}
],
"x_generator": {
"engine": "Vulnogram 1.0.2"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-35020",
"datePublished": "2026-04-06T18:58:40.513Z",
"dateRejected": "2026-05-29T16:25:15.369Z",
"dateReserved": "2026-03-31T20:40:15.618Z",
"dateUpdated": "2026-05-29T16:25:15.369Z",
"state": "REJECTED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-34452 (GCVE-0-2026-34452)
Vulnerability from nvd – Published: 2026-03-31 21:32 – Updated: 2026-04-03 16:08
VLAI
Title
Claude SDK for Python: Memory Tool Path Validation Race Condition Allows Sandbox Escape
Summary
The Claude SDK for Python provides access to the Claude API from Python applications. From version 0.86.0 to before version 0.87.0, the async local filesystem memory tool in the Anthropic Python SDK validated that model-supplied paths resolved inside the sandboxed memory directory, but then returned the unresolved path for subsequent file operations. A local attacker able to write to the memory directory could retarget a symlink between validation and use, causing reads or writes to escape the sandbox. The synchronous memory tool implementation was not affected. This issue has been patched in version 0.87.0.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: partial
CISA Coordinator (v2.0.3)
CWE
Assigner
References
3 references
| URL | Tags |
|---|---|
| https://github.com/anthropics/anthropic-sdk-pytho… | x_refsource_CONFIRM |
| https://github.com/anthropics/anthropic-sdk-pytho… | x_refsource_MISC |
| https://github.com/anthropics/anthropic-sdk-pytho… | x_refsource_MISC |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| anthropics | anthropic-sdk-python |
Affected:
>= 0.86.0, < 0.87.0
|
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-34452",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-04-03T16:06:47.039006Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-04-03T16:08:25.309Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"product": "anthropic-sdk-python",
"vendor": "anthropics",
"versions": [
{
"status": "affected",
"version": "\u003e= 0.86.0, \u003c 0.87.0"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "The Claude SDK for Python provides access to the Claude API from Python applications. From version 0.86.0 to before version 0.87.0, the async local filesystem memory tool in the Anthropic Python SDK validated that model-supplied paths resolved inside the sandboxed memory directory, but then returned the unresolved path for subsequent file operations. A local attacker able to write to the memory directory could retarget a symlink between validation and use, causing reads or writes to escape the sandbox. The synchronous memory tool implementation was not affected. This issue has been patched in version 0.87.0."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "HIGH",
"attackRequirements": "NONE",
"attackVector": "LOCAL",
"baseScore": 5.8,
"baseSeverity": "MEDIUM",
"privilegesRequired": "LOW",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:L/AC:H/AT:N/PR:L/UI:N/VC:H/VI:L/VA:N/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "NONE",
"vulnConfidentialityImpact": "HIGH",
"vulnIntegrityImpact": "LOW"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-59",
"description": "CWE-59: Improper Link Resolution Before File Access (\u0027Link Following\u0027)",
"lang": "en",
"type": "CWE"
}
]
},
{
"descriptions": [
{
"cweId": "CWE-367",
"description": "CWE-367: Time-of-check Time-of-use (TOCTOU) Race Condition",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-03-31T21:32:37.537Z",
"orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"shortName": "GitHub_M"
},
"references": [
{
"name": "https://github.com/anthropics/anthropic-sdk-python/security/advisories/GHSA-w828-4qhx-vxx3",
"tags": [
"x_refsource_CONFIRM"
],
"url": "https://github.com/anthropics/anthropic-sdk-python/security/advisories/GHSA-w828-4qhx-vxx3"
},
{
"name": "https://github.com/anthropics/anthropic-sdk-python/commit/6599043eee6e86dce16953fcd1fd828052052be6",
"tags": [
"x_refsource_MISC"
],
"url": "https://github.com/anthropics/anthropic-sdk-python/commit/6599043eee6e86dce16953fcd1fd828052052be6"
},
{
"name": "https://github.com/anthropics/anthropic-sdk-python/releases/tag/v0.87.0",
"tags": [
"x_refsource_MISC"
],
"url": "https://github.com/anthropics/anthropic-sdk-python/releases/tag/v0.87.0"
}
],
"source": {
"advisory": "GHSA-w828-4qhx-vxx3",
"discovery": "UNKNOWN"
},
"title": "Claude SDK for Python: Memory Tool Path Validation Race Condition Allows Sandbox Escape"
}
},
"cveMetadata": {
"assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"assignerShortName": "GitHub_M",
"cveId": "CVE-2026-34452",
"datePublished": "2026-03-31T21:32:37.537Z",
"dateReserved": "2026-03-27T18:18:14.895Z",
"dateUpdated": "2026-04-03T16:08:25.309Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-34451 (GCVE-0-2026-34451)
Vulnerability from nvd – Published: 2026-03-31 21:35 – Updated: 2026-04-01 18:57
VLAI
Title
Claude SDK for TypeScript: Memory Tool Path Validation Allows Sandbox Escape to Sibling Directories
Summary
Claude SDK for TypeScript provides access to the Claude API from server-side TypeScript or JavaScript applications. From version 0.79.0 to before version 0.81.0, the local filesystem memory tool in the Anthropic TypeScript SDK validated model-supplied paths using a string prefix check that did not append a trailing path separator. A model steered by prompt injection could supply a crafted path that resolved to a sibling directory sharing the memory root's name as a prefix, allowing reads and writes outside the sandboxed memory directory. This issue has been patched in version 0.81.0.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: partial
CISA Coordinator (v2.0.3)
CWE
Assigner
References
3 references
| URL | Tags |
|---|---|
| https://github.com/anthropics/anthropic-sdk-types… | x_refsource_CONFIRM |
| https://github.com/anthropics/anthropic-sdk-types… | x_refsource_MISC |
| https://github.com/anthropics/anthropic-sdk-types… | x_refsource_MISC |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| anthropics | anthropic-sdk-typescript |
Affected:
>= 0.79.0, < 0.81.0
|
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-34451",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-04-01T18:56:48.666585Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-04-01T18:57:05.442Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"product": "anthropic-sdk-typescript",
"vendor": "anthropics",
"versions": [
{
"status": "affected",
"version": "\u003e= 0.79.0, \u003c 0.81.0"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "Claude SDK for TypeScript provides access to the Claude API from server-side TypeScript or JavaScript applications. From version 0.79.0 to before version 0.81.0, the local filesystem memory tool in the Anthropic TypeScript SDK validated model-supplied paths using a string prefix check that did not append a trailing path separator. A model steered by prompt injection could supply a crafted path that resolved to a sibling directory sharing the memory root\u0027s name as a prefix, allowing reads and writes outside the sandboxed memory directory. This issue has been patched in version 0.81.0."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "HIGH",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 6.3,
"baseSeverity": "MEDIUM",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:H/AT:N/PR:N/UI:N/VC:L/VI:L/VA:N/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "NONE",
"vulnConfidentialityImpact": "LOW",
"vulnIntegrityImpact": "LOW"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-22",
"description": "CWE-22: Improper Limitation of a Pathname to a Restricted Directory (\u0027Path Traversal\u0027)",
"lang": "en",
"type": "CWE"
}
]
},
{
"descriptions": [
{
"cweId": "CWE-41",
"description": "CWE-41: Improper Resolution of Path Equivalence",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-03-31T21:35:21.018Z",
"orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"shortName": "GitHub_M"
},
"references": [
{
"name": "https://github.com/anthropics/anthropic-sdk-typescript/security/advisories/GHSA-5474-4w2j-mq4c",
"tags": [
"x_refsource_CONFIRM"
],
"url": "https://github.com/anthropics/anthropic-sdk-typescript/security/advisories/GHSA-5474-4w2j-mq4c"
},
{
"name": "https://github.com/anthropics/anthropic-sdk-typescript/commit/0ac69b3438ee9c96b21a7d3c39c07b7cdb6995d9",
"tags": [
"x_refsource_MISC"
],
"url": "https://github.com/anthropics/anthropic-sdk-typescript/commit/0ac69b3438ee9c96b21a7d3c39c07b7cdb6995d9"
},
{
"name": "https://github.com/anthropics/anthropic-sdk-typescript/releases/tag/sdk-v0.81.0",
"tags": [
"x_refsource_MISC"
],
"url": "https://github.com/anthropics/anthropic-sdk-typescript/releases/tag/sdk-v0.81.0"
}
],
"source": {
"advisory": "GHSA-5474-4w2j-mq4c",
"discovery": "UNKNOWN"
},
"title": "Claude SDK for TypeScript: Memory Tool Path Validation Allows Sandbox Escape to Sibling Directories"
}
},
"cveMetadata": {
"assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"assignerShortName": "GitHub_M",
"cveId": "CVE-2026-34451",
"datePublished": "2026-03-31T21:35:21.018Z",
"dateReserved": "2026-03-27T18:18:14.895Z",
"dateUpdated": "2026-04-01T18:57:05.442Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-34450 (GCVE-0-2026-34450)
Vulnerability from nvd – Published: 2026-03-31 21:32 – Updated: 2026-04-01 13:35
VLAI
Title
Claude SDK for Python: Insecure Default File Permissions in Local Filesystem Memory Tool
Summary
The Claude SDK for Python provides access to the Claude API from Python applications. From version 0.86.0 to before version 0.87.0, the local filesystem memory tool in the Anthropic Python SDK created memory files with mode 0o666, leaving them world-readable on systems with a standard umask and world-writable in environments with a permissive umask such as many Docker base images. A local attacker on a shared host could read persisted agent state, and in containerized deployments could modify memory files to influence subsequent model behavior. Both the synchronous and asynchronous memory tool implementations were affected. This issue has been patched in version 0.87.0.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: partial
CISA Coordinator (v2.0.3)
CWE
Assigner
References
3 references
| URL | Tags |
|---|---|
| https://github.com/anthropics/anthropic-sdk-pytho… | x_refsource_CONFIRM |
| https://github.com/anthropics/anthropic-sdk-pytho… | x_refsource_MISC |
| https://github.com/anthropics/anthropic-sdk-pytho… | x_refsource_MISC |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| anthropics | anthropic-sdk-python |
Affected:
>= 0.86.0, < 0.87.0
|
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-34450",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-04-01T13:35:12.188726Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-04-01T13:35:18.879Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"product": "anthropic-sdk-python",
"vendor": "anthropics",
"versions": [
{
"status": "affected",
"version": "\u003e= 0.86.0, \u003c 0.87.0"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "The Claude SDK for Python provides access to the Claude API from Python applications. From version 0.86.0 to before version 0.87.0, the local filesystem memory tool in the Anthropic Python SDK created memory files with mode 0o666, leaving them world-readable on systems with a standard umask and world-writable in environments with a permissive umask such as many Docker base images. A local attacker on a shared host could read persisted agent state, and in containerized deployments could modify memory files to influence subsequent model behavior. Both the synchronous and asynchronous memory tool implementations were affected. This issue has been patched in version 0.87.0."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "LOCAL",
"baseScore": 4.8,
"baseSeverity": "MEDIUM",
"privilegesRequired": "LOW",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:L/AC:L/AT:N/PR:L/UI:N/VC:L/VI:L/VA:N/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "NONE",
"vulnConfidentialityImpact": "LOW",
"vulnIntegrityImpact": "LOW"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-276",
"description": "CWE-276: Incorrect Default Permissions",
"lang": "en",
"type": "CWE"
}
]
},
{
"descriptions": [
{
"cweId": "CWE-732",
"description": "CWE-732: Incorrect Permission Assignment for Critical Resource",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-03-31T21:32:53.617Z",
"orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"shortName": "GitHub_M"
},
"references": [
{
"name": "https://github.com/anthropics/anthropic-sdk-python/security/advisories/GHSA-q5f5-3gjm-7mfm",
"tags": [
"x_refsource_CONFIRM"
],
"url": "https://github.com/anthropics/anthropic-sdk-python/security/advisories/GHSA-q5f5-3gjm-7mfm"
},
{
"name": "https://github.com/anthropics/anthropic-sdk-python/commit/715030ceb4d6dd8d3546e999c680e29532bf1255",
"tags": [
"x_refsource_MISC"
],
"url": "https://github.com/anthropics/anthropic-sdk-python/commit/715030ceb4d6dd8d3546e999c680e29532bf1255"
},
{
"name": "https://github.com/anthropics/anthropic-sdk-python/releases/tag/v0.87.0",
"tags": [
"x_refsource_MISC"
],
"url": "https://github.com/anthropics/anthropic-sdk-python/releases/tag/v0.87.0"
}
],
"source": {
"advisory": "GHSA-q5f5-3gjm-7mfm",
"discovery": "UNKNOWN"
},
"title": "Claude SDK for Python: Insecure Default File Permissions in Local Filesystem Memory Tool"
}
},
"cveMetadata": {
"assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"assignerShortName": "GitHub_M",
"cveId": "CVE-2026-34450",
"datePublished": "2026-03-31T21:32:53.617Z",
"dateReserved": "2026-03-27T18:18:14.895Z",
"dateUpdated": "2026-04-01T13:35:18.879Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-22561 (GCVE-0-2026-22561)
Vulnerability from nvd – Published: 2026-03-31 15:30 – Updated: 2026-05-10 13:58
VLAI
Summary
Uncontrolled search path elements in Anthropic Claude for Windows installer (Claude Setup.exe) versions prior to 1.1.3363 allow local privilege escalation via DLL search-order hijacking. The installer loads DLLs (e.g., profapi.dll) from its own directory after UAC elevation, enabling arbitrary code execution if a malicious DLL is planted alongside the installer.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: partial
CISA Coordinator (v2.0.3)
CWE
- CWE-427 - Uncontrolled Search Path Element
Assigner
References
1 reference
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| Anthropic | Claude Desktop - Windows |
Affected:
0 , < 1.1.3363
(semver)
|
Credits
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-22561",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-03-31T16:27:37.909291Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-427",
"description": "CWE-427 Uncontrolled Search Path Element",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-05-10T13:58:31.956Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"product": "Claude Desktop - Windows",
"vendor": "Anthropic",
"versions": [
{
"lessThan": "1.1.3363",
"status": "affected",
"version": "0",
"versionType": "semver"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "finder",
"value": "Kazuma Matsumoto, a security reseaercher at GMO Cybersecurity by IERAE, Inc"
}
],
"descriptions": [
{
"lang": "en",
"value": "Uncontrolled search path elements in Anthropic Claude for Windows installer (Claude Setup.exe) versions prior to 1.1.3363 allow local privilege escalation via DLL search-order hijacking. The installer loads DLLs (e.g., profapi.dll) from its own directory after UAC elevation, enabling arbitrary code execution if a malicious DLL is planted alongside the installer."
}
],
"metrics": [
{
"cvssV4_0": {
"baseScore": 4.7,
"baseSeverity": "MEDIUM",
"vectorString": "CVSS:4.0/AV:L/AC:L/AT:N/PR:L/UI:P/VC:N/VI:N/VA:N/SC:H/SI:H/SA:H",
"version": "4.0"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-03-31T18:13:52.328Z",
"orgId": "36234546-b8fa-4601-9d6f-f4e334aa8ea1",
"shortName": "hackerone"
},
"references": [
{
"url": "https://trust.anthropic.com/resources?s=1cvig6ldp3zvuj1yffzr11\u0026name=cve-2026-22561-dll-search-order-hijacking-in-claude-for-windows-installer"
}
]
}
},
"cveMetadata": {
"assignerOrgId": "36234546-b8fa-4601-9d6f-f4e334aa8ea1",
"assignerShortName": "hackerone",
"cveId": "CVE-2026-22561",
"datePublished": "2026-03-31T15:30:38.876Z",
"dateReserved": "2026-01-07T15:39:03.440Z",
"dateUpdated": "2026-05-10T13:58:31.956Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-33068 (GCVE-0-2026-33068)
Vulnerability from nvd – Published: 2026-03-20 08:17 – Updated: 2026-03-20 13:48
VLAI
Title
Claude Code has a Workspace Trust Dialog Bypass via Repo-Controlled Settings File
Summary
Claude Code is an agentic coding tool. Versions prior to 2.1.53 resolved the permission mode from settings files, including the repo-controlled .claude/settings.json, before determining whether to display the workspace trust confirmation dialog. A malicious repository could set permissions.defaultMode to bypassPermissions in its committed .claude/settings.json, causing the trust dialog to be silently skipped on first open. This allowed a user to be placed into a permissive mode without seeing the trust confirmation prompt, making it easier for an attacker-controlled repository to gain tool execution without explicit user consent. This issue has been patched in version 2.1.53.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: total
CISA Coordinator (v2.0.3)
CWE
- CWE-807 - Reliance on Untrusted Inputs in a Security Decision
Assigner
References
1 reference
| URL | Tags |
|---|---|
| https://github.com/anthropics/claude-code/securit… | x_refsource_CONFIRM |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| anthropics | claude-code |
Affected:
< 2.1.53
|
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-33068",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "total"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-03-20T13:48:28.263574Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-03-20T13:48:36.014Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"product": "claude-code",
"vendor": "anthropics",
"versions": [
{
"status": "affected",
"version": "\u003c 2.1.53"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "Claude Code is an agentic coding tool. Versions prior to 2.1.53 resolved the permission mode from settings files, including the repo-controlled .claude/settings.json, before determining whether to display the workspace trust confirmation dialog. A malicious repository could set permissions.defaultMode to bypassPermissions in its committed .claude/settings.json, causing the trust dialog to be silently skipped on first open. This allowed a user to be placed into a permissive mode without seeing the trust confirmation prompt, making it easier for an attacker-controlled repository to gain tool execution without explicit user consent. This issue has been patched in version 2.1.53."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "PRESENT",
"attackVector": "NETWORK",
"baseScore": 7.7,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "PASSIVE",
"vectorString": "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",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "HIGH",
"vulnIntegrityImpact": "HIGH"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-807",
"description": "CWE-807: Reliance on Untrusted Inputs in a Security Decision",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-03-20T08:17:47.794Z",
"orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"shortName": "GitHub_M"
},
"references": [
{
"name": "https://github.com/anthropics/claude-code/security/advisories/GHSA-mmgp-wc2j-qcv7",
"tags": [
"x_refsource_CONFIRM"
],
"url": "https://github.com/anthropics/claude-code/security/advisories/GHSA-mmgp-wc2j-qcv7"
}
],
"source": {
"advisory": "GHSA-mmgp-wc2j-qcv7",
"discovery": "UNKNOWN"
},
"title": "Claude Code has a Workspace Trust Dialog Bypass via Repo-Controlled Settings File"
}
},
"cveMetadata": {
"assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
"assignerShortName": "GitHub_M",
"cveId": "CVE-2026-33068",
"datePublished": "2026-03-20T08:17:47.794Z",
"dateReserved": "2026-03-17T19:27:06.343Z",
"dateUpdated": "2026-03-20T13:48:36.014Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}