What is the CVSS score of CVE-2026-30635?

CVE-2026-30635 has a CVSS 3.1 base score of 8.1 (High). CVSS vector: CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H. EPSS exploitation probability: 0.2%.

Which versions of automagik-genie are affected by CVE-2026-30635?

automagik-genie 2.5.27 MCP Server contains a remote code execution vulnerability affecting systems that process transcripts from external endpoints, allowing attackers to execute arbitrary commands…

How to fix or mitigate CVE-2026-30635?

Within 24 hours: Identify all automagik-genie 2.5.27 deployments and document which systems process external or untrusted transcript sources; isolate affected servers from production networks if possible. Within 7 days: Implement network controls to restrict FORGE_BASE_URL endpoint access to known-trusted sources only; monitor for suspicious transcript processing activity. Within 30 days: Upgrade to automagik-genie 2.5.28 or later once vendor releases patched version; if upgrade unavailable, evaluate alternative MCP server solutions and discontinue use of 2.5.27 in untrusted data processing workflows.

automagik-genie CVE-2026-30635

EUVD-2026-29159 HIGH

OS Command Injection (CWE-78)

2026-05-11 mitre

GHSA-64vr-4gr2-m642

Command Injection

8.1

CVSS 3.1 · NVD

Severity by source

NVD PRIMARY

8.1 HIGH

AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H

Primary rating from NVD · only source for this CVE.

CVSS VectorNVD

CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H

Attack Vector

Network

Attack Complexity

High

Privileges Required

None

User Interaction

None

Scope

Unchanged

Confidentiality

High

Integrity

High

Availability

High

Lifecycle Timeline

Analysis Generated

May 12, 2026 - 16:31 vuln.today

CVSS changed

May 12, 2026 - 14:22 NVD

8.1 (HIGH)

CVE Published

May 11, 2026 - 00:00 nvd

UNKNOWN (no severity yet)

CVE Published

May 11, 2026 - 00:00 nvd

HIGH 8.1

DescriptionCVE.org

Command injection vulnerability in automagik-genie 2.5.27 MCP Server allows attackers to execute arbitrary commands via the view_task (aka view) in the readTranscriptFromCommit function in dist/mcp/server.js when a user reads from an external FORGE_BASE_URL.

AnalysisAI

Remote code execution in automagik-genie 2.5.27 MCP Server occurs when processing transcripts from attacker-controlled external FORGE_BASE_URL endpoints. Exploitation chains command injection in the readTranscriptFromCommit function's view_task parameter to execute arbitrary system commands on the server. A proof-of-concept exploit exists, though active exploitation has not been confirmed by CISA KEV at time of analysis.

Technical ContextAI

This vulnerability exploits CWE-78 (OS Command Injection) in the Model Context Protocol (MCP) Server component of automagik-genie. The readTranscriptFromCommit function in dist/mcp/server.js improperly sanitizes user input from the view_task (aliased as 'view') parameter before passing it to system command execution. When the application fetches transcript data from an external FORGE_BASE_URL controlled by an attacker, malicious command sequences embedded in the response can break out of the intended command context and execute arbitrary OS commands with the privileges of the MCP Server process. The vulnerability exists specifically in version 2.5.27, affecting the server-side transcript processing pipeline.

RemediationAI

Primary remediation requires upgrading automagik-genie beyond version 2.5.27 to a patched release, though the specific fixed version is not confirmed in available data as no vendor security bulletin was identified. Until a verified patch is released, implement these compensating controls: (1) Disable or restrict the readTranscriptFromCommit functionality if not business-critical, accepting loss of external transcript reading capability. (2) Implement strict allowlist validation for FORGE_BASE_URL to permit only trusted, organization-controlled endpoints, recognizing this reduces integration flexibility with third-party forges. (3) Deploy Web Application Firewall (WAF) rules to inspect and block command injection patterns in view_task parameters, though this may generate false positives requiring tuning. (4) Run the MCP Server process under least-privilege service accounts within restricted containers or sandboxes to limit command execution blast radius, with operational overhead of container management. Monitor the researcher disclosure at https://gist.github.com/spdc-elm/3ddecd10ffa85c5963ab7fe531619875 and the project's official repository for patch announcements.