registry / llm-wiki-loop / 1.3.1
llm-wiki-loop@1.3.1
The production framework for self-improving & self-organizing LLM knowledge vaults — grounding invariants, code drift detection, auto-skillification, and 1-click multi-agent setup.
AI Security Review
scanned 22d ago · by lpm-firewall-aiAutomatic deployment and replacement of an agent extension creates a concrete lifecycle risk. The writes are restricted to the package-owned wiki-manager skill, supporting a warning rather than a malicious verdict.
Static reason
High-risk behavior combination matched malicious policy.; source matched previously finalized malicious package; routed for review
Trigger
npm postinstall; explicit CLI install or init commands also deploy the extension.
Impact
Installation or later package updates can replace wiki-manager instructions loaded by an agent runtime. The inspected behavior remains confined to the package-owned extension.
Mechanism
The installer detects runtime directories and recursively copies bundled skill files with overwrite enabled, without requesting consent.
AI rationale
Automatic deployment and replacement of an agent extension creates a concrete lifecycle risk. The writes are restricted to the package-owned wiki-manager skill, supporting a warning rather than a malicious verdict.
Evidence
package.jsonscripts/install.jsbin/cli.js.claude.config/opencode.agents.cursor.gemini.commandcode.windsurfwiki-manager
Decision evidence
public snapshotAI called this Suspicious at 96.0% confidence as Dangerous Capability with low false-positive risk.
Evidence for warning
- package.json automatically runs scripts/install.js after installation.
- scripts/install.js recursively copies the bundled wiki-manager extension and overwrites existing files in its destination.
- The installer targets detected global or project agent skill directories without a consent prompt.
- bin/cli.js also exposes skill installation and vault initialization through explicit user commands.
Evidence against
- Installer writes are confined to the package-owned wiki-manager extension rather than broad agent configuration.
- No credential export, remote payload execution, obfuscation, or runtime self-dependency was identified in the inspected entrypoints.
Behavioral surface
ChildProcessDynamicRequireEnvironmentVarsFilesystem
HighEntropyStringsUrlStrings
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 unpkgbin/cli.jsView file
7L8: const fs = require('fs');
L9: const os = require('os');
Medium
Dynamic Require
Package source references dynamic require/import behavior.
bin/cli.jsView on unpkg · L7scripts/install.js
Findings
1 Critical3 High5 Medium4 Low
CriticalAi Agent Control Hijackscripts/install.js
HighInstall Time Lifecycle Scriptspackage.json
HighKnown Malware Source Similarityscripts/install.js
HighKnown Malware Source Similarityscripts/scan-skill-candidates.js
MediumAmbiguous Install Lifecycle Scriptpackage.json
MediumDynamic Requirebin/cli.js
MediumEnvironment Vars
MediumShips Build Helperskills/wiki-manager/scripts/check_evidence.py
MediumStructural Risk Force Deep Review
LowScripts Present
LowFilesystem
LowHigh Entropy Strings
LowUrl Strings
Affected versions and remediation
This report applies to llm-wiki-loop@1.3.1.
See all versions of llm-wiki-loop for other recorded verdicts.
- Review the evidence and your use of llm-wiki-loop@1.3.1 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.