CVE-2026-58195
Received Received - Intake

Command Injection in Agentic-Flow MCP Server

Publication date: 2026-07-17

Last updated on: 2026-07-17

Assigner: GitHub, Inc.

Description
Agentic-Flow is an AI agent orchestration platform. Prior to 2.0.14, agentic-flow MCP server tools in src/mcp/standalone-stdio.ts, src/mcp/fastmcp/servers/claude-flow-sdk.ts, src/mcp/fastmcp/servers/stdio-full.ts, src/mcp/fastmcp/servers/http-streaming-updated.ts, src/mcp/fastmcp/servers/http-sse.ts, src/mcp/fastmcp/servers/poc-stdio.ts, src/mcp/fastmcp/tools/agent/{execute,list,parallel}.ts, src/mcp/fastmcp/tools/swarm/orchestrate.ts, and src/mcp/fastmcp/tools/hooks/pretrain.ts interpolated attacker-influenceable tool parameters such as agent, task, name, language, and agentdb directly into shell command strings passed to execSync(), allowing arbitrary OS command execution with the privileges of the MCP server user. This issue is fixed in version 2.0.14.

CVSS Scores

EPSS Scores

Probability:
Percentile:

Meta Information

Published
2026-07-17
Last Modified
2026-07-17
Generated
2026-07-18
AI Q&A
2026-07-17
EPSS Evaluated
N/A
NVD
CVE-2026-58195
EUVD
EUVD-2026-45256

Affected Vendors & Products

Showing 4 associated CPEs
Vendor Product Version / Range
agentic-flow agentic-flow to 2.0.14 (exc)
ruvnet agentic-flow to 2.0.14 (exc)
ruvnet agentic-flow From 2.0.14 (exc)
ruvnet ruflo From 3.12.4 (inc)

Helpful Resources

Exploitability

CWE
CWE Icon
KEV
KEV Icon
CWE ID Description
CWE-78 The product constructs all or part of an OS command using externally-influenced input from an upstream component, but it does not neutralize or incorrectly neutralizes special elements that could modify the intended OS command when it is sent to a downstream component.

Attack-Flow Graph

AI Quick Actions

Instant insights powered by AI
Executive Summary

This is an OS command injection vulnerability (CWE-78) in the agentic-flow MCP server tools. Attacker-controlled parameters like agent, task, or name are directly interpolated into shell command strings passed to execSync(). This allows arbitrary OS command execution with the privileges of the MCP server user. The flaw exists in multiple MCP server files and was fixed in version 2.0.14 by replacing execSync() with execFileSync() using argv arrays and shell: false.

Detection Guidance

Check for vulnerable agentic-flow versions by running: npm list agentic-flow. If version is <=2.0.13, the system is vulnerable. Inspect node_modules/agentic-flow/dist/mcp/standalone-stdio.js for execSync calls. Use grep -r "execSync(" in the agentic-flow directory to find vulnerable patterns.

Impact Analysis

An attacker could exploit this to execute arbitrary commands on your system with the privileges of the user running the MCP server. This is possible even without direct access to the host if the AI agent processes untrusted content. The HTTP/SSE transports increase risk by exposing vulnerable sinks without authentication. Impact includes data theft, system compromise, or denial of service.

Compliance Impact

This vulnerability could lead to unauthorized data access, modification, or exfiltration, violating GDPR's integrity and confidentiality requirements and HIPAA's security rules for protected health information. Organizations using affected versions may face compliance violations, regulatory penalties, and loss of trust due to potential data breaches.

Mitigation Strategies

Upgrade agentic-flow to version 2.0.14 or later using npm install agentic-flow@^2.0.14. For downstream packages like ruflo, run npm install ruflo@latest. Verify the fix by checking for absence of execSync calls in node_modules/agentic-flow/dist/mcp/standalone-stdio.js.

Chat Assistant

Ask questions about this CVE
Hi! I’m here to help you understand CVE-2026-58195. Ask me anything about the vulnerability, its impact, or mitigation strategies.
0/70

EPSS Chart