registry / aims-workstation / 0.1.12
aims-workstation@0.1.12
AIMS 工作站：本地智能体工作站——ARM 助手本地化（下载/离线/升级）、工作台对话式定制、多工作区隔离、分级自主与审计
AI Security Review
scanned 10d ago · by lpm-firewall-aiThe scanner's credential warning is not supported by the inspected source: ARM credentials are used for ARM API operations and sent to the configured ARM service. The LLM endpoint and key are also exposed as workstation configuration with a documented default; no concrete attack behavior was established.
Static reason
One or more suspicious static signals were detected.; previous stored version diff introduced dangerous source
Trigger
The operator uses an ARM integration feature, such as publishing or downloading a UI asset.
Impact
The inspected transfer supports the workstation's service integration; source inspection found no unrelated credential recipient or confirmed attack surface.
Mechanism
The server adds the configured ARM token as a bearer credential to requests sent to the configured ARM base, which has a documented service default.
AI rationale
The scanner's credential warning is not supported by the inspected source: ARM credentials are used for ARM API operations and sent to the configured ARM service. The LLM endpoint and key are also exposed as workstation configuration with a documented default; no concrete attack behavior was established.
Evidence
package.jsonserver.jsweb/dist/assets/index-Cicx_wyb.js.env.example
Network endpoints2
arm.agent-platform.dev.aimstek.cnbifrost.infra.dev.aimstek.cn/v1
Decision evidence
public snapshotAI called this Clean at 91.0% confidence as Benign with low false-positive risk.
Evidence for AI clean decision
- The manifest exposes a workstation command and has no install lifecycle hook; the inspected dependency is an LLM API library.
- ARM credentials come from workstation settings, environment variables, or the local ARM config, and the destination is a configurable ARM base with a documented service default.
- ARM upload and download calls send the ARM token as a bearer credential to that selected base during API requests.
- The inspected callers use those requests for UI asset publishing, downloading, and installation.
- The remaining runtime operation invokes the panel checker and records its warnings for audit.
- The settings interface allows the operator to set the ARM base and token; the environment example documents the ARM and LLM service defaults.
Evidence against
- The credential transfers inspected are consistent with the workstation's ARM integration; no source evidence showed hidden forwarding to an unrelated recipient.
Behavioral surface
ChildProcessCryptoEnvironmentVarsEval
Source & flagged code
6 flagged.env.exampleView file
19patternName = npm_token
severity = critical
line = 19
Critical
19patternName = npm_token
severity = critical
line = 19
Critical
scripts/check-panel.mjsView file
20try {
L21: const factory = new Function('React', 'html', 'ui', 'useState', 'useEffect', 'useRef', 'useMemo',
L22: `"use strict";\n${code}\n;if (typeof Panel !== 'function') throw new Error('no Panel');\nreturn Panel;`);
Low
Eval
Package source references a known benign dynamic code generation pattern.
scripts/check-panel.mjsView on unpkg · L20Findings
2 Critical2 High4 Medium6 Low
CriticalCritical Secret.env.example
CriticalSecret Pattern.env.example
HighCredential Redirect Persistenceserver.js
HighPrevious Version Dangerous Deltaserver.js
MediumNetwork
MediumEnvironment Vars
MediumShips Build Helperscripts/publish-npm.sh
MediumStructural Risk Force Deep Review
LowScripts Present
LowEvalscripts/check-panel.mjs
LowFilesystem
LowHigh Entropy Strings
LowUrl Strings
LowNo License
Affected versions and remediation
This report applies to aims-workstation@0.1.12.
See all versions of aims-workstation for other recorded verdicts.
- Review the evidence and your use of aims-workstation@0.1.12 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.