@melaya/runner@1.1.30
Run Melaya AI pipelines locally with your own LM Studio or Ollama models
AI Security Review
scanned 2mo ago · by lpm-firewall-aiThe package is a high-risk remote agent runner with credential-access and remote-code execution capabilities. Static inspection found no install-time malicious chain or covert exfiltration beyond its service-integrated runtime behavior.
Static reason
One or more suspicious static signals were detected.; previous stored version diff introduced dangerous source
Trigger
The user runs melaya-runner with a runner token and the service sends a run or connector request.
Impact
A compromised or abusive service can execute code in the user's runner context and receive connector credentials.
Mechanism
Remote bundle installation, server-dispatched code execution, and credential forwarding.
AI rationale
The package is a high-risk remote agent runner with credential-access and remote-code execution capabilities. Static inspection found no install-time malicious chain or covert exfiltration beyond its service-integrated runtime behavior.
Evidence
dist/detect.jsdist/sharedVendor.jsdist/connection.jsdist/pythonEnv.jsdist/linkedinLogin.js~/.melaya-runner/shared/~/.melaya-runner/agentscope/~/.melaya-runner/venv/main.py
Network endpoints2
api.melaya.org/api/v1/runner/shared-bundlewww.linkedin.com
Decision evidence
public snapshotAI called this Suspicious at 94.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- It reads existing Claude, Codex, and GitHub authentication material during local model detection.
- It downloads a server-controlled Python bundle and writes its supplied files into the user's Melaya cache.
- It verifies and then executes Python supplied in a server socket payload.
- It captures LinkedIn session cookies and returns them to the connected server.
- It creates a Python environment and automatically installs a large dependency set, including browser-stealth tooling.
Evidence against
- The manifest has no install, preinstall, or postinstall hook; its only lifecycle entry is prepublishOnly.
- The sensitive and remote-execution paths require running the CLI and connecting it with a runner token.
- The inspected browser authorization code blocks private and cloud-metadata destinations rather than accessing them.
Behavioral surface
ChildProcessCryptoEnvironmentVarsFilesystemNetworkShellWebSocket
Source & flagged code
8 flaggeddist/codeWorker.jsView file
51// model (--experimental-permission / --permission): fs read/write
L52: // limited to the scratch dir, child_process and worker_threads denied.
L53: //
High
Child Process
Package source references child process execution.
dist/codeWorker.jsView on unpkg · L51288// ulimit -t: hard CPU time in seconds.
L289: // We use /bin/sh -c "ulimit ...; exec sandbox-exec ..." so both limits
L290: // apply before the Node process starts. This is a soft-then-hard pair;
High
dist/browserProvisioner.jsView file
•matchType = previous_version_dangerous_delta
matchedPackage = @melaya/runner@1.0.93
matchedIdentity = npm:QG1lbGF5YS9ydW5uZXI:1.0.93
similarity = 0.538
summary = stored previous version shares package body but lacks this dangerous source file
Findings
1 Critical6 High4 Medium6 Low
CriticalPrevious Version Dangerous Deltadist/browserProvisioner.js
HighChild Processdist/codeWorker.js
HighShelldist/codeWorker.js
HighSandbox Evasion Gated Capabilitydist/browserProvisioner.js
HighCloud Metadata Accessdist/browserAuthz.js
HighRuntime Package Installdist/lumaLogin.js
HighSemantic Analysis Limiteddist/linkedinLogin.js#virtual:normalized:round1
MediumNetwork
MediumEnvironment Vars
MediumShips Build Helperdist/localRagIngest.py
MediumStructural Risk Force Deep Review
LowNon Install Lifecycle Scripts
LowScripts Present
Affected versions and remediation
This report applies to @melaya/runner@1.1.30.
See all versions of @melaya/runner for other recorded verdicts.
- Review the evidence and your use of @melaya/runner@1.1.30 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.