registry / @lion1ou/skill-ctl / 1.2.2
@lion1ou/skill-ctl@1.2.2
Configuration-driven Agent Skill catalog installer.
AI Security Review
scanned 2d ago · by lpm-firewall-aiThe flagged process execution supports user-requested external skill preparation, while the manager binds to loopback and protects API access with a random token. Default dashboard analytics is disclosed, limited as documented, and has an opt-out, so the reviewed behavior is consistent with the package’s purpose.
Static reason
One or more suspicious static signals were detected.; previous stored version diff introduced dangerous source
Trigger
User runs the CLI; dashboard telemetry starts when the user runs manage and opens its pages.
Impact
The reviewed behavior supports the package’s stated purpose; disclosed analytics reports anonymous usage and is documented as excluding skill content, paths, and the access token.
Mechanism
Selected external skills are prepared through npx; the dashboard may load an analytics script from track.lion1ou.tech unless telemetry is disabled.
AI rationale
The flagged process execution supports user-requested external skill preparation, while the manager binds to loopback and protects API access with a random token. Default dashboard analytics is disclosed, limited as documented, and has an opt-out, so the reviewed behavior is consistent with the package’s purpose.
Network endpoints2
skills.sh/api/searchtrack.lion1ou.tech
Decision evidence
public snapshotAI called this Clean at 94.0% confidence as Benign with low false-positive risk.
Evidence for AI clean decision
- The package has no install or postinstall hook; its lifecycle scripts are build, test, prepack, and prepublishOnly.
- The dashboard telemetry is limited to anonymous page views, frontend errors, and performance, and can be disabled with SKILLCTL_NO_TELEMETRY.
- The README discloses the telemetry, says it excludes skill content, paths, and the access token, and documents the opt-out.
- The manager server binds to 127.0.0.1 and uses a random per-process token for API protection.
- Catalog installation runs after the user invokes install and selects skills and targets.
- Search requests go to skills.sh, the service used to find external skills.
Evidence against
- The dashboard loads analytics from track.lion1ou.tech by default when the user runs the management UI.
- Installing selected external skills can invoke npx to prepare them.
Behavioral surface
ChildProcessCryptoEnvironmentVarsFilesystemNetwork
Source & flagged code
4 flaggeddist/dev-session.jsView file
1import { execFileSync, spawn } from 'node:child_process';
L2: import { readFileSync } from 'node:fs';
High
Child Process
Package source references child process execution.
dist/dev-session.jsView on unpkg · L1dist/core/skills-source.jsView file
21export function inspectPackageSkills(packageId) {
L22: const result = spawnSync('npx', ['-y', 'skills', 'add', packageId, '-l'], {
L23: encoding: 'utf8',
High
Runtime Package Install
Package source invokes a package manager install command at runtime.
dist/core/skills-source.jsView on unpkg · L21dist/core/workspace-picker.js#virtual:normalized:round1View file
Findings
1 Critical3 High3 Medium5 Low
CriticalPrevious Version Dangerous Deltadist/core/catalog-install.js
HighChild Processdist/dev-session.js
HighRuntime Package Installdist/core/skills-source.js
HighSemantic Analysis Limiteddist/core/workspace-picker.js#virtual:normalized:round1
MediumNetwork
MediumEnvironment Vars
MediumStructural Risk Force Deep Review
LowNon Install Lifecycle Scripts
LowScripts Present
LowFilesystem
LowHigh Entropy Strings
LowUrl Strings
Affected versions and remediation
This report applies to @lion1ou/skill-ctl@1.2.2.
See all versions of @lion1ou/skill-ctl for other recorded verdicts.
- Review the evidence and your use of @lion1ou/skill-ctl@1.2.2 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.