registry / @taichuai/taichu-cli / 1.0.6
@taichuai/taichu-cli@1.0.6
TaiChu CLI - Manage TaiChu models, usage, and configuration from your terminal
AI Security Review
scanned 28d ago · by lpm-firewall-aiThe inspected CLI entrypoint uses package-aligned TaiChu endpoints and rejects redirects; the scanner’s suspicious signals do not establish an attack. Machine fingerprinting is present, but the inspected behavior is consistent with host identification and does not show credential exfiltration.
Static reason
One or more suspicious static signals were detected.
Trigger
Running the CLI or invoking its authentication and service features.
Impact
No concrete unauthorized credential disclosure or other attack was established; machine identification has privacy implications.
Mechanism
The program uses system information and commands to derive a host identifier, while its configured service endpoints use TaiChu domains.
AI rationale
The inspected CLI entrypoint uses package-aligned TaiChu endpoints and rejects redirects; the scanner’s suspicious signals do not establish an attack. Machine fingerprinting is present, but the inspected behavior is consistent with host identification and does not show credential exfiltration.
Evidence
package.jsondist/bin/taichu.js
Network endpoints2
cloud.zidongtaichu.com/taichu-web/auth.zidongtaichu.com
Decision evidence
public snapshotAI called this Clean at 83.0% confidence as Benign with low false-positive risk.
Evidence for AI clean decision
- The manifest exposes the bundled CLI as the package executable and runs a build during prepare.
- The bundled CLI uses TaiChu cloud and authentication endpoints for its service requests.
- The request path sends requests to their supplied URL and rejects redirects.
- With no CLI arguments, the program starts its interactive REPL; the inspected source shows no automatic credential forwarding to an unrelated endpoint.
Evidence against
- The CLI gathers machine identifiers through system commands and persists a host ID, consistent with device identification but privacy-sensitive.
Behavioral surface
ChildProcessCryptoEnvironmentVarsFilesystemNetworkShell
HighEntropyStringsMinifiedPossibleObfuscationTrivialUrlStrings
Source & flagged code
7 flaggeddist/bin/taichu.jsView file
1#!/usr/bin/env node
L2: import e,{dirname as t,join as n,resolve as r}from"path";import"url";import{homedir as o,platform as a,machine as i}from"os";import s,{existsSync as l,readFileSync as u,mkdirSync a...
High
1#!/usr/bin/env node
L2: import e,{dirname as t,join as n,resolve as r}from"path";import"url";import{homedir as o,platform as a,machine as i}from"os";import s,{existsSync as l,readFileSync as u,mkdirSync a...
High
1#!/usr/bin/env node
L2: import e,{dirname as t,join as n,resolve as r}from"path";import"url";import{homedir as o,platform as a,machine as i}from"os";import s,{existsSync as l,readFileSync as u,mkdirSync a...
High
Same File Env Network Execution
A single source file combines environment access, network access, and code or shell execution; review context before blocking.
dist/bin/taichu.jsView on unpkg · L1Findings
7 High3 Medium5 Low
HighChild Processdist/bin/taichu.js
HighShelldist/bin/taichu.js
HighSame File Env Network Executiondist/bin/taichu.js
HighCredential Redirect Persistencedist/bin/taichu.js
HighCommand Output Exfiltrationdist/bin/taichu.js
HighSandbox Evasion Gated Capabilitydist/bin/taichu.js
HighTrigger Reachable Command Output Exfiltrationdist/bin/taichu.js
MediumNetwork
MediumEnvironment Vars
MediumStructural Risk Force Deep Review
LowNon Install Lifecycle Scripts
LowScripts Present
LowFilesystem
LowHigh Entropy Strings
Affected versions and remediation
This report applies to @taichuai/taichu-cli@1.0.6.
See all versions of @taichuai/taichu-cli for other recorded verdicts.
- Review the evidence and your use of @taichuai/taichu-cli@1.0.6 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.