registry / @melaya/runner / 1.1.79
@melaya/runner@1.1.79
Run Melaya AI pipelines locally with your own LM Studio or Ollama models
AI Security Review
scanned 6h ago · by lpm-firewall-aiThe flagged credential flows query Anthropic and GitHub Copilot using their respective provider credentials to discover available models. Startup diagnostics use the user-supplied runner token and send bounded fields to the Melaya service; no malicious behavior was found.
Static reason
One or more suspicious static signals were detected.; source matched previously finalized malicious package; routed for review
Trigger
Provider discovery runs when the user starts the CLI; failure reporting runs on selected startup failures when a runner token is supplied.
Impact
No unrelated credential exfiltration, cloud metadata collection, or other malicious behavior was identified.
Mechanism
The scanner reads provider login state and queries the corresponding provider APIs. Failure diagnostics go to the Melaya service and contain bounded status and platform details.
AI rationale
The flagged credential flows query Anthropic and GitHub Copilot using their respective provider credentials to discover available models. Startup diagnostics use the user-supplied runner token and send bounded fields to the Melaya service; no malicious behavior was found.
Evidence
dist/cli.jsdist/detect.jsdist/installReport.jsdist/browserAuthz.js
Network endpoints4
api.anthropic.comapi.github.comapi.githubcopilot.comapi.melaya.org
Decision evidence
public snapshotAI called this Clean at 94.0% confidence as Benign with low false-positive risk.
Evidence for AI clean decision
- The CLI starts a provider scan when the runner is launched.
- Claude OAuth is used to query Anthropic’s model catalog at api.anthropic.com.
- A GitHub token is sent to GitHub’s Copilot token endpoint.
- Install-failure reporting sends bounded diagnostic fields to the Melaya service using the runner token supplied by the user.
- Browser authorization code blocks cloud metadata and private network addresses.
Evidence against
Behavioral surface
ChildProcessCryptoDynamicRequireEnvironmentVarsFilesystemNetworkShellWebSocket
HighEntropyStringsUrlStrings
Source & flagged code
20 flaggeddist/uvBootstrap.jsView file
20import { homedir, platform, arch } from "os";
L21: import { execFile } from "child_process";
L22: import { spawnTracked as spawn, registerSetupChild, terminateTree } from "./processTree.js";
High
Child Process
Package source references child process execution.
dist/uvBootstrap.jsView on unpkg · L20•matchType = normalized_sha256
matchedPackage = @melaya/runner@1.1.73
matchedPath = dist/uvBootstrap.js
matchedIdentity = npm:QG1lbGF5YS9ydW5uZXI:1.1.73
similarity = 1.000
summary = normalized source hash matched finalized malicious source
High
Known Malware Source Similarity
Source file is highly similar to a previously finalized malicious package; route for source-aware review.
dist/uvBootstrap.jsView on unpkgdist/lumaLogin.jsView file
52const child = spawn("npx", ["playwright", "install", "chromium"], {
L53: shell: true,
L54: stdio: ["ignore", "pipe", "pipe"],
Findings
18 High5 Medium5 Low
HighChild Processdist/uvBootstrap.js
HighShelldist/lumaLogin.js
HighSandbox Evasion Gated Capabilitydist/browserProvisioner.js
HighCloud Metadata Accessdist/browserAuthz.js
HighRuntime Package Installdist/lumaLogin.js
HighSemantic Analysis Limiteddist/cli.js#virtual:normalized:round1
HighKnown Malware Source Similaritydist/browserProvisioner.js
HighKnown Malware Source Similaritydist/codeWorker.js
HighKnown Malware Source Similaritydist/detect.js
HighKnown Malware Source Similaritydist/vaults.js
HighKnown Malware Source Similaritydist/cli.js
High
Affected versions and remediation
This report applies to @melaya/runner@1.1.79.
See all versions of @melaya/runner for other recorded verdicts.
- Review the evidence and your use of @melaya/runner@1.1.79 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.