ux-atlas@0.1.7
UX skills for agents. Each skill is skills/<name>/SKILL.md.
AI Security Review
scanned 5d ago · by lpm-firewall-aiThe package automatically modifies project agent configuration during postinstall, creating a residual lifecycle risk. The changes point to package-owned instructions and are limited to marked blocks, so the inspected behavior does not establish a publish-blocking attack.
Static reason
High-risk behavior combination matched malicious policy.
Trigger
A local npm installation runs the postinstall hook.
Impact
Agents may load the package's instructions automatically after installation; no separate malicious behavior was established in the inspected source.
Mechanism
The initializer adds or updates marked pointers in AGENTS.md and CLAUDE.md that direct agents to ux-atlas instructions and skills, then creates missing .atlas configuration files.
AI rationale
The package automatically modifies project agent configuration during postinstall, creating a residual lifecycle risk. The changes point to package-owned instructions and are limited to marked blocks, so the inspected behavior does not establish a publish-blocking attack.
Evidence
package.jsonscripts/init.mjsAGENTS.mdCLAUDE.md.atlas/config.json.atlas/README.md
Decision evidence
public snapshotAI called this Suspicious at 92.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- The package manifest runs scripts/init.mjs from its postinstall hook.
- The initializer targets the project root from INIT_CWD and skips only global installs or paths outside node_modules; it has no consent prompt.
- The install path adds or updates marked pointers in the project's AGENTS.md and CLAUDE.md.
- Those pointers direct agents to this package's own instructions and skills.
- The initializer creates .atlas configuration and a guide only when those files are absent.
Evidence against
- The agent configuration changes are limited to marked package pointers, and the setup points to ux-atlas's own files.
- The inspected installer and scanner show no network endpoint or credential transfer.
Behavioral surface
CryptoEnvironmentVarsFilesystem
HighEntropyStrings
Source & flagged code
7 flagged · loading sourcepackage.jsonView file
•scripts.postinstall = node scripts/init.mjs
High
Install Time Lifecycle Scripts
Package defines install-time lifecycle scripts.
package.jsonView on unpkg24"scripts": {
"lint": "eslint .",
"postinstall": "node scripts/init.mjs",
"index": "node scripts/
Medium
Suspicious Lifecycle Evidence
The package manifest runs scripts/init.mjs from its postinstall hook.
package.jsonView on unpkg · L24scripts/scan.mjsView file
105const content = fs.readFileSync(file);
L106: const text = content.toString('utf8');
L107: if (isTokens) {
Low
Weak Crypto
Package source references weak cryptographic algorithms.
scripts/scan.mjsView on unpkg · L105Findings
1 Critical2 High5 Medium4 Low
CriticalAi Agent Control Hijackscripts/init.mjs
HighInstall Time Lifecycle Scriptspackage.json
HighExternal Ai Agent Control Surface Mutationscripts/init.mjs
MediumEnvironment Vars
MediumStructural Risk Force Deep Review
MediumSuspicious Lifecycle Evidencepackage.json
MediumAi Review Evidencescripts/init.mjs
MediumAi Review Evidencescripts/init.mjs
LowScripts Present
LowWeak Cryptoscripts/scan.mjs
LowFilesystem
LowHigh Entropy Strings
Affected versions and remediation
This report applies to ux-atlas@0.1.7.
See all versions of ux-atlas for other recorded verdicts.
- Review the evidence and your use of ux-atlas@0.1.7 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.