CVE-2026-54051: Network-AI: Improper Neutralization of Special Elements used in an OS Command

Should you act?

Critical, preconditions apply

Severe if exploited, but exploitation needs conditions that may not apply to you. This signal is computed from CISA KEV, EPSS and CVSS data; it is not an editorial verdict.

• Not in CISA KEV, but CISA’s SSVC assessment records a public proof-of-concept, so attack code exists.
• EPSS 0.7%, percentile 51: FIRST’s model estimates a 0.7% chance of exploitation activity in the next 30 days, higher than 51% of all scored CVEs (moderate). EPSS is a probability, not a verdict.
• Needs an account (low (any account)). Exposure depends on who can log in.
• Only applies if your dependency tree includes network-ai (npm) < 5.9.1, fixed in 5.9.1. Check your lockfile, including transitive dependencies.

What it is

Network-AI is a TypeScript/Node.js multi-agent orchestrator. Prior to version 5.9.1, the agent sandbox gates shell commands behind an allowlist (SandboxPolicy.isCommandAllowed), which THREATMODEL.md calls the main control against a compromised agent (Adversary 3.2). The allowlist glob-matches the whole command string, but ShellExecutor runs that string through /bin/sh -c. So any wildcard allow such as git , npm or node also matches git status; , and a scoped command becomes arbitrary execution. The issue is fixed in v5.9.1. ShellExecutor now executes via spawn(file, args, { shell: false }) using a quote-aware parsed argv, so no shell is invoked. SandboxPolicy.isCommandAllowed and the new SandboxPolicy.tokenizeCommand reject any unquoted shell metacharacter (; & | $ ( ) { } newline) or unterminated quote before the allowlist glob match; quoted metacharacters are preserved as literal argument data. Users should upgrade to network-ai@5.9.1 or later. As defense in depth, avoid broad wildcard allowlist entries such as node / npm which are direct code execution by design.

From the GitHub advisory

Summary

The agent sandbox gates shell commands behind an allowlist (SandboxPolicy.isCommandAllowed), which THREATMODEL.md calls the main control against a compromised agent (Adversary 3.2). The allowlist glob-matches the whole command string, but ShellExecutor runs that string through /bin/sh -c. So any wildcard allow such as git , npm or node also matches git status; , and a scoped command becomes arbitrary execution.

Root cause

Matching and execution disagree on what a command is. Lines pinned to 40e42d7 (lib/agent-runtime.ts is identical to the v5.8.5 tag).

1. isCommandAllowed matches the full string, with no tokenizing and no metacharacter check:

Weakness type: CWE-78.

Product: network-ai (npm) · Assigned by: GitHub_M · Published: 2026-07-20

Affected and fixed versions

Severity scores

• Attack vector: network (remotely reachable)
• Attack complexity: low
• Privileges required: low (any account)
• User interaction: none
• Scope: changed (can affect other components)

CVSS measures how bad a flaw is if exploited, not how likely exploitation is. Read it next to the exploitation data below.

Exploitation and urgency

• CISA KEV: not listed (no confirmed in-the-wild exploitation known to CISA).
• CISA SSVC assessment: exploitation poc, automatable no, technical impact total.
• FIRST EPSS (model of 2026-10-21): 0.67% probability of exploitation activity in the next 30 days; percentile 51.

Timeline

• 2026-06-19 — GitHub-reviewed advisory GHSA-qw6v-5fcf-5666 published (critical)
• 2026-07-20 — CVE record published by GitHub_M
• 2026-07-21 — CVE record last updated

Sources and references

Original sources first. Version and fix details should always be confirmed against the vendor or project advisory.

Vulnerability records

1. GitHub Advisory GHSA-qw6v-5fcf-5666 github.com
2. CVE.org record CVE-2026-54051 cve.org
3. NVD entry CVE-2026-54051 nvd.nist.gov

Other references from the CVE record and advisory

1. github.com github.com
2. github.com github.com

Critical npm advisories · Today’s Security Digest