@playlist-tech/enterprise-skills@1.5.26-enterprise.1
An enterprise-customizable fork of Vercel Labs' skills CLI for internal skill registries.
AI Security Review
scanned 2mo ago · by lpm-firewall-aiThe package has a real explicit-user AI-agent control-surface capability, so it warrants a warning. Static inspection found no unconsented install-time execution, embedded command payload, credential theft, or self-reinstall chain.
Static reason
One or more suspicious static signals were detected.
Trigger
The user runs the CLI setup command with SKILLS_HOOK_STOP_CMD or SKILLS_HOOK_FAIL_CMD set.
Impact
Subsequent agent events may run those configured commands; remote skills can also be copied or linked into agent skill directories.
Mechanism
Writes environment-supplied commands into AI-agent hook settings.
AI rationale
The package has a real explicit-user AI-agent control-surface capability, so it warrants a warning. Static inspection found no unconsented install-time execution, embedded command payload, credential theft, or self-reinstall chain.
Evidence
dist/cli.mjs.claude/settings.json
Network endpoints2
add-skill.vercel.sh/tadd-skill.vercel.sh/audit
Decision evidence
public snapshotAI called this Suspicious at 92.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- The explicit setup command writes environment-provided commands into detected AI-agent hook configuration files.
- The CLI installs selected skill content into agent skill directories by copy or symlink.
- The CLI sends opt-out telemetry and audit requests to add-skill.vercel.sh.
Evidence against
- package.json has no preinstall, install, or postinstall hook; prepare only invokes husky.
- The executable only imports the CLI after a user runs its bin command.
- Hook wiring requires an explicit setup command and nonempty SKILLS_HOOK_* environment values.
- No runtime dependency on this package itself was found.
Behavioral surface
ChildProcessCryptoEnvironmentVarsFilesystemNetworkShell
HighEntropyStringsUrlStrings
WildcardDependency
Source & flagged code
3 flaggeddist/_chunks/libs/simple-git.mjsView file
4import { Buffer as Buffer$1 } from "node:buffer";
L5: import { spawn } from "child_process";
L6: import { normalize } from "node:path";
High
Child Process
Package source references child process execution.
dist/_chunks/libs/simple-git.mjsView on unpkg · L4dist/cli.mjsView file
•Manifest-trigger-reachable source links an external AI-agent control path to a behavior-bearing write operation.
;
}
function isvalidrelativepath(path) {
return path.startswith("./");
}
async function getpluginskillpaths(basepath) {
const searchdirs = [];
const addpluginskillpaths = (pluginbase, skills) => {
if (!iscontainedin(pluginbase, basepath)) return;
if (skills && skills.length > 0) for (const skillpath of skills) {
if (!isvalidrelativepath(skillpath)) continue;
const skilldir = dirname(join(pluginbase, skillpath));
if (iscontainedin(skilldir, basepath)) searchdirs.push(skilldir);
}
searchdirs.push(join(pluginbase, "skills"));
};
try {
const content = await readfile(join(basepath, ".claude-plugin/marketplace.json"), "utf-8");
const manifest = json.parse(con
High
Findings
4 High3 Medium5 Low
HighChild Processdist/_chunks/libs/simple-git.mjs
HighShell
HighTrigger Reachable External Ai Agent Control Surface Mutationdist/cli.mjs
HighCross File Remote Execution Contextdist/cli.mjs
MediumNetwork
MediumEnvironment Vars
MediumWildcard Dependency
LowNon Install Lifecycle Scripts
LowScripts Present
LowFilesystem
LowHigh Entropy Strings
LowUrl Strings
Affected versions and remediation
This report applies to @playlist-tech/enterprise-skills@1.5.26-enterprise.1.
See all versions of @playlist-tech/enterprise-skills for other recorded verdicts.
- Review the evidence and your use of @playlist-tech/enterprise-skills@1.5.26-enterprise.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.