registry / Install hook abuse / litclaude-ai
litclaude-ai
Claude Code-native workflow distribution.
This report shows the evidence for litclaude-ai@0.4.3. 3 of 15 known versions are blocked or flagged. The newest version, 0.4.6, is flagged. See all versions.
AI Security Review
scanned 2mo ago · by lpm-firewall-aiThis is an unconsented postinstall mutation of a broad AI-agent control surface, combined with a runtime same-package update chain. The source establishes the complete install-hook abuse path even though it does not show secret theft.
Static reason
One or more suspicious static signals were detected.; source matched previously finalized malicious package; routed for review; previous stored version diff introduced dangerous source
Trigger
Global npm installation triggers postinstall; later interactive install, update, or doctor commands can trigger the updater.
Impact
The package can persist a plugin and alter Claude Code permissions, registry, marketplace, HUD, and settings without an explicit setup command.
Mechanism
Automatic Claude configuration mutation and same-package version-selected npm execution.
Attack narrative
On a global npm install, postinstall invokes the package's install command. That command copies and enables a Claude plugin, changes the Claude registry and marketplace, and updates global settings. The package also maintains an update mechanism that runs npm with a selected release of litclaude-ai and invokes that release's install command, creating a nested lifecycle path that can replace the Claude control surface.
AI rationale
This is an unconsented postinstall mutation of a broad AI-agent control surface, combined with a runtime same-package update chain. The source establishes the complete install-hook abuse path even though it does not show secret theft.
Evidence
package.jsonscripts/postinstall.mjsbin/litclaude-ai.jsplugins/litclaude/lib/automatic-update.mjs~/.claude/plugins/cache/litclaude-ai~/.claude/plugins/installed_plugins.json~/.claude/plugins/known_marketplaces.json~/.claude/settings.json~/.litclaude
Network endpoints1
registry.npmjs.org/litclaude-ai/latest
Decision evidence
public snapshotAI called this Malicious at 99.0% confidence as Malware with low false-positive risk.
Evidence for block
- The package runs a postinstall hook that invokes its install command during global npm installation.
- That install command copies a plugin, enables it in Claude Code, and changes global Claude settings.
- Its automatic updater executes npm with a version-selected release of the same package and then runs its install command.
- The updater backs up and replaces Claude plugin registry, marketplace, and settings files.
Evidence against
- The postinstall hook skips ordinary non-global installs, CI, and source checkouts unless forced.
- The registry request is limited to the npm registry and the updater sanitizes its child environment.
- No source evidence shows credential collection or exfiltration.
Behavioral surface
ChildProcessCryptoDynamicRequireEnvironmentVarsFilesystemNetworkShell
HighEntropyStringsUrlStrings
Source & flagged code
16 flagged · loading sourcepackage.jsonView file
•scripts.postinstall = node scripts/postinstall.mjs
High
Install Time Lifecycle Scripts
Package defines install-time lifecycle scripts.
package.jsonView on unpkg•scripts.postinstall = node scripts/postinstall.mjs
Medium
Ambiguous Install Lifecycle Script
Install-time lifecycle script is not statically allowlisted and needs review.
package.jsonView on unpkgscripts/qa-uiux-visual-qa-scenarios.mjsView file
83const fixtureBytes = await bounded(fixturesRoot, "scenarios.json");
L84: const strict = await import(pathToFileURL(join(installedRoot, "lib/strict-json.mjs")).href);
L85: const tui = await import(pathToFileURL(
Medium
Dynamic Require
Package source references dynamic require/import behavior.
scripts/qa-uiux-visual-qa-scenarios.mjsView on unpkg · L83Findings
13 High6 Medium4 Low
HighInstall Time Lifecycle Scriptspackage.json
HighPayload In Excluded Dirplugins/litclaude/vendor/045_scientific-visualization/tests/test_figure_export.py
HighSemantic Analysis Limitedbin/litclaude-ai.js#virtual:normalized:round1
HighKnown Malware Source Similarityplugins/litclaude/lib/automatic-update.mjs
HighKnown Malware Source Similaritybin/update-notifier.mjs
HighKnown Malware Source Similarityplugins/litclaude/skills/lsp-setup/scripts/verify-lsp.ts
HighKnown Malware Source Similarityscripts/qa-negative-gate-matrix.mjs
HighKnown Malware Source Similarityplugins/litclaude/skills/programming/scripts/typescript/check-no-excuse-rules.ts
HighKnown Malware Source Similarityscripts/inspect-agent-tools.mjs
Affected versions and remediation
This report applies to litclaude-ai@0.4.3.
The versions list below shows the verdict for each recorded version.
- Avoid installing litclaude-ai@0.4.3. Remove it from direct dependencies and check your lockfile for transitive copies.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
- If this version ran, investigate the affected machine and build environment. Rotate credentials it could access and rebuild from a trusted environment.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.
Versions
15 known versions0.4.6latestSuspiciousAgent extension lifecycle riskWarnPublishedSep 3, 2026, 4:10 PM0.4.3shownMaliciousInstall Hook AbuseBlockPublishedAug 30, 2026, 9:35 PM0.3.46MaliciousSoft block: AI-agent control surfaceBlockPublishedAug 15, 2026, 11:19 PM0.3.44MaliciousSoft block: AI-agent control surfaceBlockPublishedAug 14, 2026, 5:18 AM