@melaya/runner@1.1.54
⚠ Under reviewRun Melaya AI pipelines locally with your own LM Studio or Ollama models
AI Security Review
scanned 12d ago · by lpm-firewall-aiThe inspected credential-discovery callers and runner connection behavior do not establish a concrete attack. The Luma browser bridge remains unresolved after bounded inspection, so the source evidence does not support a clean verdict.
Decision evidence
public snapshot- The manifest exposes a command-line runner and uses prepublishOnly only to build the package.
- Provider detection reads saved Claude and Codex credentials and invokes provider detection from the CLI; the inspected callers do not show those values being sent to an unrelated recipient.
- The connection code passes tool and approval settings into the runner environment, and browser authorization code contains private-address classification.
- The inspected browser bridge describes loading saved Luma browser session state; its full behavior remains unresolved.
- Static analysis reported an alias-growth limit in the Luma browser bridge, and the inspected excerpts do not resolve its complete request flow or recipients.
Source & flagged code
8 flaggedPackage source references child process execution.
dist/uvBootstrap.jsView on unpkg · L20Findings
Affected versions and remediation
This report applies to @melaya/runner@1.1.54.
See all versions of @melaya/runner for other recorded verdicts.
- Review the evidence and your use of @melaya/runner@1.1.54 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.