registry / @mnemonik/cli / 7.138.1
@mnemonik/cli@7.138.1
Mnemonik installer and local project helper
AI Security Review
scanned 22d ago · by lpm-firewall-aiThe package’s explicit setup flow can modify selected user-level agent integrations using first-party host packages, which warrants a warning under the agent integration policy. The npm lifecycle, scanner hint, and dynamic evaluation findings did not establish a confirmed attack.
Static reason
One or more suspicious static signals were detected.
Trigger
A user runs the CLI install command and proceeds through its setup flow.
Impact
Selected agent configuration may be changed to enable Mnemonik integrations.
Mechanism
The CLI selects host targets, loads their adapter, plans configuration changes, and stages and applies those changes through its install journal.
AI rationale
The package’s explicit setup flow can modify selected user-level agent integrations using first-party host packages, which warrants a warning under the agent integration policy. The npm lifecycle, scanner hint, and dynamic evaluation findings did not establish a confirmed attack.
Evidence
package.jsondist/router.jsdist/install/journey.jsdist/install/hosts.js
Decision evidence
public snapshotAI called this Suspicious at 86.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- package.json declares no lifecycle scripts and lists Mnemonik-owned host hook packages.
- The explicit install command runs a setup flow with a host selection checklist; non-interactive setup requires acceptance and apply flags.
- For selected user-level targets, host adapters plan configuration changes that the install journal stages and applies.
- The agent-configuration search hints land on repeated scanner ignore comments that suppress only the Claude worktree directory.
- The eval match is in bundled WebAssembly parser glue; the CLI updater checks release signatures and package integrity.
Evidence against
- npm installation does not trigger the host setup; the package has no lifecycle scripts.
- Host integration changes require a user-invoked CLI setup flow.
- The scanner hints do not show agent configuration writes, and the inspected source did not establish credential theft or other malicious behavior.
Behavioral surface
ChildProcessCryptoDynamicRequireEnvironmentVars
Source & flagged code
7 flaggeddist/diagnostics.jsView file
898import { createHash, createPublicKey, randomUUID, timingSafeEqual, verify as verifySignature } from "node:crypto";
L899: import { execFile, execFileSync } from "node:child_process";
L900: function windowsCommand(file) {
High
Child Process
Package source references child process execution.
dist/diagnostics.jsView on unpkg · L8985213var func = `(${args}) => { ${body} };`;
L5214: ASM_CONSTS[start] = eval(func);
L5215: }
High
1import { createRequire as __createRequire } from "node:module"; const require = __createRequire(import.meta.url);
L2: var __create = Object.create;
Medium
Dynamic Require
Package source references dynamic require/import behavior.
dist/diagnostics.jsView on unpkg · L1Findings
1 Critical5 High4 Medium4 Low
CriticalRemote Response Code Executiondist/runtime/selfUpdate.js
HighChild Processdist/diagnostics.js
HighEvaldist/diagnostics.js
HighSame File Env Network Executiondist/install/journey.js
HighSandbox Evasion Gated Capabilitydist/install/journey.js
HighCross File Remote Execution Contextdist/auth/credentials.js
MediumDynamic Requiredist/diagnostics.js
MediumNetwork
MediumEnvironment Vars
MediumStructural Risk Force Deep Review
LowFilesystem
LowHigh Entropy Strings
LowUrl Strings
LowNo License
Affected versions and remediation
This report applies to @mnemonik/cli@7.138.1.
See all versions of @mnemonik/cli for other recorded verdicts.
- Review the evidence and your use of @mnemonik/cli@7.138.1 before allowing it. Restrict the permissions described in this report.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.