registry / agenthub-multiagent-mcp / 1.75.0
agenthub-multiagent-mcp@1.75.0
MCP server for AgentHub multi-agent communication
AI Security Review
scanned 4d ago · by lpm-firewall-aiThe inspected network and credential flows support the package’s stated AgentHub worker function. Persistence-like shell changes require an explicit setup command, and worker settings are written to a temporary file.
Static reason
No blocking static signals were detected.
Trigger
The worker connects when a user runs its CLI; shell setup runs only when explicitly requested.
Impact
Inspected behavior supports the package’s multi-agent workflow; no unrelated credential transfer or automatic profile persistence was found.
Mechanism
The worker sends its AgentHub API key to the configured AgentHub service and launches Claude with its authentication environment. Session capture settings are stored in a temporary file.
AI rationale
The inspected network and credential flows support the package’s stated AgentHub worker function. Persistence-like shell changes require an explicit setup command, and worker settings are written to a temporary file.
Evidence
~/.agenthub/worker.json~/.agenthub/worker.log$TMPDIR/agenthub-worker-settings-<pid>.json
Network endpoints1
agenthub.contetial.com
Decision evidence
public snapshotAI called this Clean at 91.0% confidence as Benign with low false-positive risk.
Evidence for AI clean decision
- The manifest exposes explicit CLI entry points and has no install, preinstall, or postinstall hook.
- The worker authenticates to AgentHub using its configured API key, consistent with the package’s stated purpose.
- The Claude OAuth token is retained in the environment passed to the spawned Claude process; the inspected code does not send it to AgentHub.
- Worker capture settings are written to a temporary file for a launched session.
- Shell profile changes are performed only when the setup command receives the explicit install-shell subcommand.
Evidence against
Behavioral surface
ChildProcessCryptoDynamicRequireEnvironmentVarsFilesystemNativeBindingsNetworkShellWebSocket
HighEntropyStringsUrlStrings
Source & flagged code
7 flaggednative/agenthub-memvid/index.jsView file
6L7: const { existsSync, readFileSync } = require('fs')
L8: const { join } = require('path')
Medium
Dynamic Require
Package source references dynamic require/import behavior.
native/agenthub-memvid/index.jsView on unpkg · L6dist/commands/setup-shell.jsView file
25Trigger-reachable persistence chain: manifest.bin -> dist/setup.js -> dist/commands/setup-shell.js
L25: export function detectShell() {
L26: const shellPath = process.env.SHELL;
L27: if (!shellPath)
...
L35: export function rcFilePath(shell) {
L36: const home = os.homedir();
L37: return path.join(home, shell === "bash" ? ".bashrc" : ".zshrc");
L38: }
High
Trigger Reachable Persistence
A manifest entrypoint or package-local install chain reaches persistence behavior.
dist/commands/setup-shell.jsView on unpkg · L2525
Findings
3 High7 Medium4 Low
HighTrigger Reachable External Ai Agent Control Surface Mutationdist/worker.js
HighTrigger Reachable Persistencedist/commands/setup-shell.js
HighSemantic Analysis Limiteddist/channelSetup.js#virtual:normalized:round1
MediumDynamic Requirenative/agenthub-memvid/index.js
MediumNetwork
MediumEnvironment Vars
MediumInstall Persistencedist/commands/setup-shell.js
MediumShips Native Binarynative/agenthub-memvid/agenthub-memvid.win32-x64-msvc.node
MediumShips Build Helperdist/run-claude.sh
MediumStructural Risk Force Deep Review
LowScripts Present
LowFilesystem
LowHigh Entropy Strings
Affected versions and remediation
This report applies to agenthub-multiagent-mcp@1.75.0.
See all versions of agenthub-multiagent-mcp for other recorded verdicts.
- Review the evidence and your use of agenthub-multiagent-mcp@1.75.0 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.