@yuandc/aica@0.1.54
AICA remote Agent worker
AI Security Review
scanned 25d ago · by lpm-firewall-aiThe flagged behavior is consistent with the active remote worker implementation and requires explicit CLI activation; the platform gate is ordinary path normalization. Inspection found no automatic install attack, hidden recipient, or concrete malicious chain.
Decision evidence
public snapshot- package.json defines start and help scripts, with no automatic install lifecycle or runtime self-dependency.
- dist/core/aca-config.js uses platform checks to normalize project paths; they do not conceal execution.
- dist/core/job-worker.js requires a configured token before claiming remote jobs, and dist/commands/start.js starts the worker loops through an explicit CLI command.
- dist/core/aca-server-client.js disables TLS verification only through the explicit ACA_SERVER_INSECURE_TLS environment option.
- The Codex launcher configures a dedicated provider key and disables OpenAI authentication for that provider.
- dist/core/file-transfer-worker.js resolves real paths and rejects file requests outside the selected project root.
- The running worker accepts remote agent jobs and filesystem operations, giving the authenticated server substantial local access.
- Explicitly enabling insecure TLS weakens transport security.
Source & flagged code
4 flaggedSource gates dangerous network, credential, or execution behavior behind CI, host, platform, time, or geo fingerprint checks.
dist/core/aca-config.jsView on unpkg · L1dist/core/aca-config.js uses platform checks to normalize project paths; they do not conceal execution.
dist/core/aca-config.jsView on unpkg · L1Findings
Affected versions and remediation
This report applies to @yuandc/aica@0.1.54.
See all versions of @yuandc/aica for other recorded verdicts.
- Review the evidence and your use of @yuandc/aica@0.1.54 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.