registry / @sokeai/cli / 1.0.116
@sokeai/cli@1.0.116
授客AI官方CLI工具 - 支持AI Agent Skills
AI Security Review
scanned 2d ago · by lpm-firewall-aiThe postinstall hook automatically installs a bundled skill into detected agent skill directories without asking, leaving a first-party agent-extension lifecycle risk. The inspected skill includes authentication and write-confirmation safeguards, and the source does not establish broader control-surface abuse or another concrete attack.
Static reason
High-risk behavior combination matched malicious policy.; source matched previously finalized malicious package; routed for review
Trigger
Installing the package with npm, which runs its postinstall hook.
Impact
The skill becomes available to those agents for portal-management requests. The inspected skill requires an authenticated organization check and confirmation before write operations; no broader agent-setting change was identified.
Mechanism
After downloading the binary, the installer detects existing agent directories and copies the portal skill into each detected skills directory.
AI rationale
The postinstall hook automatically installs a bundled skill into detected agent skill directories without asking, leaving a first-party agent-extension lifecycle risk. The inspected skill includes authentication and write-confirmation safeguards, and the source does not establish broader control-surface abuse or another concrete attack.
Evidence
package.jsonscripts/install.jsskills/multi/soke-portal-management/SKILL.md
Network endpoints2
soke-frontendstatic-cdn.soke.cn/soke-cli/releasesgithub.com/sokeai/soke-cli/releases/download
Decision evidence
public snapshotAI called this Suspicious at 94.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- The manifest runs scripts/install.js automatically through postinstall.
- The installer copies the bundled portal skill into detected agent skill directories, including Claude’s, without asking during installation.
- The copied skill gives an agent access to the package’s portal-management commands when users request that task.
- The installer fetches the CLI binary from the package’s CDN and GitHub release endpoints.
- The launcher passes command-line arguments to the downloaded CLI binary.
Evidence against
- The installer copies one bundled, first-party skill; inspected source does not show it changing agent settings or enabling private prompt, response, or tool-content export.
- The skill directs portal operations through the CLI, requires an authenticated current-organization check, and requires confirmation before write operations.
Behavioral surface
ChildProcessCryptoEnvironmentVarsFilesystemNetworkShell
Source & flagged code
7 flagged · loading sourcepackage.jsonView file
•scripts.postinstall = node scripts/install.js
High
Install Time Lifecycle Scripts
Package defines install-time lifecycle scripts.
package.jsonView on unpkg•scripts.postinstall = node scripts/install.js
Medium
Ambiguous Install Lifecycle Script
Install-time lifecycle script is not statically allowlisted and needs review.
package.jsonView on unpkgscripts/install.jsView file
14if (!fs.existsSync(srcDir)) return;
L15: if (!fs.existsSync(destDir)) fs.mkdirSync(destDir, { recursive: true });
L16:
...
L37:
L38: fs.copyFileSync(srcPath, destPath);
L39: }
...
L50: { root: path.join(homeDir, '.workbuddy'), skills: path.join(homeDir, '.workbuddy', 'skills') },
L51: { root: path.join(homeDir, '.claude'), skills: path.join(homeDir, '.claude', 'skills') },
L52: { root: path.join(homeDir, '.workclaw'), skills: path.join(homeDir, '.workclaw', 'skills') },
...
L116: const config = JSON.parse(configText);
L117: const workspaceDir = config?.defaults?.agents?.openaiAgents?.main?.workspace;
L118: if (typeof workspaceDir === 'string' && workspaceDir.length > 0) {
Findings
1 Critical4 High5 Medium4 Low
CriticalAi Agent Control Hijackscripts/install.js
HighInstall Time Lifecycle Scriptspackage.json
HighInstall Named Payload Filescripts/install.js
HighSemantic Analysis Limitedscripts/run.js
HighKnown Malware Source Similarityscripts/run.js
MediumAmbiguous Install Lifecycle Scriptpackage.json
MediumNetwork
MediumEnvironment Vars
MediumShips Build Helperskills/multi/soke-ai-coach-director/scripts/gate_check.py
MediumStructural Risk Force Deep Review
LowScripts Present
LowFilesystem
LowHigh Entropy Strings
Low
Affected versions and remediation
This report applies to @sokeai/cli@1.0.116.
See all versions of @sokeai/cli for other recorded verdicts.
- Review the evidence and your use of @sokeai/cli@1.0.116 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.