CVE-2026-30310

{"dataType": "CVE_RECORD", "cveMetadata": {"state": "PUBLISHED", "cveId": "CVE-2026-30310", "assignerOrgId": "8254265b-2729-46b6-b9e3-3dfca2d5bfca", "assignerShortName": "mitre", "dateUpdated": "2026-04-01T18:39:56.827Z", "dateReserved": "2026-03-04T00:00:00.000Z", "datePublished": "2026-03-31T00:00:00.000Z"}, "containers": {"cna": {"providerMetadata": {"orgId": "8254265b-2729-46b6-b9e3-3dfca2d5bfca", "shortName": "mitre", "dateUpdated": "2026-03-31T13:32:25.751Z"}, "descriptions": [{"lang": "en", "value": "In its design for automatic terminal command execution, Sixth offers two options: Execute safe commands and Execute all commands. The description for the former states that commands determined by the model to be safe will be automatically executed, whereas if the model judges a command to be potentially destructive, it still requires user approval. However, this design is highly susceptible to prompt injection attacks. An attacker can employ a generic template to wrap any malicious command and mislead the model into misclassifying it as a 'safe' command, thereby bypassing the user approval requirement and resulting in arbitrary command execution."}], "affected": [{"vendor": "n/a", "product": "n/a", "versions": [{"version": "n/a", "status": "affected"}]}], "references": [{"url": "https://trysixth.com/"}, {"url": "https://github.com/Secsys-FDU/LLM-Tool-Calling-CVEs/issues/12"}], "problemTypes": [{"descriptions": [{"type": "text", "lang": "en", "description": "n/a"}]}]}, "adp": [{"problemTypes": [{"descriptions": [{"type": "CWE", "cweId": "CWE-77", "lang": "en", "description": "CWE-77 Improper Neutralization of Special Elements used in a Command ('Command Injection')"}]}], "metrics": [{"cvssV3_1": {"scope": "UNCHANGED", "version": "3.1", "baseScore": 9.8, "attackVector": "NETWORK", "baseSeverity": "CRITICAL", "vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H", "integrityImpact": "HIGH", "userInteraction": "NONE", "attackComplexity": "LOW", "availabilityImpact": "HIGH", "privilegesRequired": "NONE", "confidentialityImpact": "HIGH"}}, {"other": {"type": "ssvc", "content": {"timestamp": "2026-04-01T18:39:07.654899Z", "id": "CVE-2026-30310", "options": [{"Exploitation": "none"}, {"Automatable": "yes"}, {"Technical Impact": "total"}], "role": "CISA Coordinator", "version": "2.0.3"}}}], "title": "CISA ADP Vulnrichment", "providerMetadata": {"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0", "shortName": "CISA-ADP", "dateUpdated": "2026-04-01T18:39:56.827Z"}}]}, "dataVersion": "5.2"}

{"dataType": "CVE_RECORD", "containers": {"adp": [{"title": "CISA ADP Vulnrichment", "metrics": [{"cvssV3_1": {"scope": "UNCHANGED", "version": "3.1", "baseScore": 9.8, "attackVector": "NETWORK", "baseSeverity": "CRITICAL", "vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H", "integrityImpact": "HIGH", "userInteraction": "NONE", "attackComplexity": "LOW", "availabilityImpact": "HIGH", "privilegesRequired": "NONE", "confidentialityImpact": "HIGH"}}, {"other": {"type": "ssvc", "content": {"id": "CVE-2026-30310", "role": "CISA Coordinator", "options": [{"Exploitation": "none"}, {"Automatable": "yes"}, {"Technical Impact": "total"}], "version": "2.0.3", "timestamp": "2026-04-01T18:39:07.654899Z"}}}], "problemTypes": [{"descriptions": [{"lang": "en", "type": "CWE", "cweId": "CWE-77", "description": "CWE-77 Improper Neutralization of Special Elements used in a Command ('Command Injection')"}]}], "providerMetadata": {"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0", "shortName": "CISA-ADP", "dateUpdated": "2026-04-01T18:39:52.985Z"}}], "cna": {"affected": [{"vendor": "n/a", "product": "n/a", "versions": [{"status": "affected", "version": "n/a"}]}], "references": [{"url": "https://trysixth.com/"}, {"url": "https://github.com/Secsys-FDU/LLM-Tool-Calling-CVEs/issues/12"}], "descriptions": [{"lang": "en", "value": "In its design for automatic terminal command execution, Sixth offers two options: Execute safe commands and Execute all commands. The description for the former states that commands determined by the model to be safe will be automatically executed, whereas if the model judges a command to be potentially destructive, it still requires user approval. However, this design is highly susceptible to prompt injection attacks. An attacker can employ a generic template to wrap any malicious command and mislead the model into misclassifying it as a 'safe' command, thereby bypassing the user approval requirement and resulting in arbitrary command execution."}], "problemTypes": [{"descriptions": [{"lang": "en", "type": "text", "description": "n/a"}]}], "providerMetadata": {"orgId": "8254265b-2729-46b6-b9e3-3dfca2d5bfca", "shortName": "mitre", "dateUpdated": "2026-03-31T13:32:25.751Z"}}}, "cveMetadata": {"cveId": "CVE-2026-30310", "state": "PUBLISHED", "dateUpdated": "2026-04-01T18:39:56.827Z", "dateReserved": "2026-03-04T00:00:00.000Z", "assignerOrgId": "8254265b-2729-46b6-b9e3-3dfca2d5bfca", "datePublished": "2026-03-31T00:00:00.000Z", "assignerShortName": "mitre"}, "dataVersion": "5.2"}

{"affected": [{"affectedData": [{"product": "n/a", "vendor": "n/a", "versions": [{"status": "affected", "version": "n/a"}]}], "source": "cve@mitre.org"}], "cveTags": [], "descriptions": [{"lang": "en", "value": "In its design for automatic terminal command execution, Sixth offers two options: Execute safe commands and Execute all commands. The description for the former states that commands determined by the model to be safe will be automatically executed, whereas if the model judges a command to be potentially destructive, it still requires user approval. However, this design is highly susceptible to prompt injection attacks. An attacker can employ a generic template to wrap any malicious command and mislead the model into misclassifying it as a 'safe' command, thereby bypassing the user approval requirement and resulting in arbitrary command execution."}, {"lang": "es", "value": "En su dise\u00f1o para la ejecuci\u00f3n autom\u00e1tica de comandos de terminal, Sixth ofrece dos opciones: Ejecutar comandos seguros y Ejecutar todos los comandos. La descripci\u00f3n de la primera establece que los comandos determinados por el modelo como seguros se ejecutar\u00e1n autom\u00e1ticamente, mientras que si el modelo juzga un comando como potencialmente destructivo, a\u00fan requiere la aprobaci\u00f3n del usuario. Sin embargo, este dise\u00f1o es altamente susceptible a ataques de inyecci\u00f3n de prompts. Un atacante puede emplear una plantilla gen\u00e9rica para envolver cualquier comando malicioso y enga\u00f1ar al modelo para que lo clasifique err\u00f3neamente como un comando 'seguro', eludiendo as\u00ed el requisito de aprobaci\u00f3n del usuario y resultando en la ejecuci\u00f3n arbitraria de comandos."}], "id": "CVE-2026-30310", "lastModified": "2026-07-25T10:10:00.167", "metrics": {"cvssMetricV31": [{"cvssData": {"attackComplexity": "LOW", "attackVector": "NETWORK", "availabilityImpact": "HIGH", "baseScore": 9.8, "baseSeverity": "CRITICAL", "confidentialityImpact": "HIGH", "integrityImpact": "HIGH", "privilegesRequired": "NONE", "scope": "UNCHANGED", "userInteraction": "NONE", "vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H", "version": "3.1"}, "exploitabilityScore": 3.9, "impactScore": 5.9, "source": "134c704f-9b21-4f2e-91b3-4a467353bcc0", "type": "Secondary"}], "ssvcV203": [{"source": "134c704f-9b21-4f2e-91b3-4a467353bcc0", "ssvcData": {"id": "CVE-2026-30310", "options": [{"exploitation": "none"}, {"automatable": "yes"}, {"technicalImpact": "total"}], "role": "CISA Coordinator", "timestamp": "2026-04-01T18:39:07.654899Z", "version": "2.0.3"}}]}, "published": "2026-03-31T14:16:11.390", "references": [{"source": "cve@mitre.org", "url": "https://github.com/Secsys-FDU/LLM-Tool-Calling-CVEs/issues/12"}, {"source": "cve@mitre.org", "url": "https://trysixth.com/"}], "sourceIdentifier": "cve@mitre.org", "vulnStatus": "Awaiting Analysis", "weaknesses": [{"description": [{"lang": "en", "value": "CWE-77"}], "source": "134c704f-9b21-4f2e-91b3-4a467353bcc0", "type": "Secondary"}]}

{}

{"affected": [{"configurations": [{"platform": null, "status": "affected", "versions": null}], "enrichment": {"confidence": 80.0, "confidence_source": "inferred", "scores": [{"score": 80.0, "source": "inferred"}]}, "original": {"product": "n/a", "source": "cna", "vendor": "n/a"}, "product": "sixth", "vendor": "trysixth"}], "analysis": {"en": {"generated_at": "2026-04-02T23:04:49.285465+00:00", "value": {"mitigation_remediation": ["Disable automatic command execution in Sixth and enforce manual approval for all commands", "Apply a vendor patch or update for Sixth as soon as it becomes available", "Restrict the model\u2019s input space to eliminate untrusted prompts or implement a prompt sanitization layer", "Apply the principle of least privilege to the process executing commands from Sixth", "Monitor vendor advisories for updates or additional mitigation guidance"], "summary": {"action": "Patch ASAP", "impact": "Remote Code Execution"}, "threat_synthesis": {"affected_systems": "The issue impacts installations of the tool Sixth that have the automatic command execution feature enabled. No specific vendor, product, or version details are provided, so any deployment of Sixth carrying this feature is potentially exposed.", "description_and_impact": "The flaw lies in the way the LLM tool named Sixth classifies commands as safe or unsafe before execution. The design allows the model to decide whether a command is safe, and only requires user approval for potentially destructive commands. Experienced attackers can inject prompts that trick the model into misclassifying a malicious instruction as safe. This misclassification bypasses the user approval gate and causes the tool to execute the command automatically. The vulnerability is an instance of OS Command Injection (CWE\u201177) and can grant an attacker unrestricted ability to run arbitrary commands on the host system, jeopardizing confidentiality, integrity, and availability.", "risk_and_exploitability": "The vulnerability has a CVSS score of 9.8, indicating a very high severity. The EPSS score is below 1%, suggesting low current exploitation probability, yet the flaw remains practical for attackers skilled in prompt injection. The vulnerability is not listed in the CISA KEV catalog. Exploitation requires an attacker to supply a crafted prompt that influences the model\u2019s safety classification logic, after which arbitrary shell commands are executed with the tool\u2019s runtime privileges."}}}}, "created": "2026-03-31T19:57:00.942012+00:00", "title": "LLM Prompt Injection Misclassifies Safe Commands Leading to Arbitrary Command Execution", "updated": "2026-04-03T09:19:38.052255+00:00", "vendors": ["trysixth", "trysixth$PRODUCT$sixth"]}