keyring-agent-core@0.3.21
Core AI chat agent with multi-agent architecture, ReAct pattern, and modular tool system powered by Gemini
AI Security Review
scanned 9d ago · by lpm-firewall-aiSource inspection found no concrete attack behavior. The scanner matches correspond to package functionality with user action or configuration gates, and its cloud-metadata matches are part of a private-network denylist.
Decision evidence
public snapshot- package.json has no install hook; its prepublishOnly script runs the build.
- The native-send tool collects a recipient and amount, then asks the user to review and submit the form.
- Vector search is selected only when enabled; its types document Upstash as the backend and disabled behavior as the legacy matcher.
- Automatic knowledge-base ingestion requires autoIngest to be explicitly true.
- The data-URI path parses JSON to extract an image URL; it does not execute the decoded content.
- The eval-enabled worker runs bundled serialization code, and the metadata addresses appear in a private-network denylist.
Source & flagged code
6 flagged · loading sourceSource uses private key material to transfer cryptocurrency funds.
dist/index.native.jsView on unpkg · L1Source fetches a remote non-code asset, decodes its contents, and dynamically executes the decoded payload.
dist/index.native.jsView on unpkg · L1A single source file combines environment access, network access, and code or shell execution; review context before blocking.
dist/index.native.jsView on unpkg · L813Source reaches cloud instance metadata or link-local credential endpoints.
dist/index.native.jsView on unpkg · L1Source contains an obfuscated payload loader that reconstructs and executes hidden code.
dist/index.native.jsView on unpkg · L1A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.
dist/index.native.jsView on unpkgFindings
Affected versions and remediation
This report applies to keyring-agent-core@0.3.21.
See all versions of keyring-agent-core for other recorded verdicts.
- Review the evidence and your use of keyring-agent-core@0.3.21 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.