registry / agenthub-multiagent-mcp / 1.66.0
agenthub-multiagent-mcp@1.66.0
MCP server for AgentHub multi-agent communication
AI Security Review
scanned 16d ago · by lpm-firewall-aiThe package automatically adds a first-party inbox hook to global Claude settings during MCP startup. Its prompt includes limits on replies, but automatic agent communication before the user's first request remains a concrete capability risk.
Static reason
No blocking static signals were detected.
Trigger
Starting the MCP server, unless the inbox hook is disabled through its environment setting.
Impact
The agent may communicate based on session context without a direct user prompt, subject to the reminder's stated limits.
Mechanism
The package writes a SessionStart command to the user's Claude settings; that command prompts the agent to read inbox messages and reply.
AI rationale
The package automatically adds a first-party inbox hook to global Claude settings during MCP startup. Its prompt includes limits on replies, but automatic agent communication before the user's first request remains a concrete capability risk.
Evidence
dist/index.jsdist/inboxHookInstall.jsdist/inboxHookCli.js~/.claude/settings.json
Decision evidence
public snapshotAI called this Suspicious at 82.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- The manifest lists build, development, start, lint, typecheck, and test scripts.
- MCP startup installs the inbox hook, and the hook is enabled by default unless its environment setting is disabled.
- The installer adds a SessionStart command to the user's global Claude settings and writes the updated file.
- The reminder directs the agent to check messages and reply before handling the user's first request.
- The reminder tells the agent to escalate secrets, personal data, decisions, and requests with side effects.
- The cited worker locations are local log-stream writes.
Evidence against
- The inbox reminder limits replies involving secrets, personal data, user decisions, or side effects.
- The worker citations resolve to logging behavior, not an additional attack operation.
Behavioral surface
ChildProcessCryptoEnvironmentVarsFilesystemNativeBindingsNetworkShellWebSocket
Source & flagged code
6 flaggeddist/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 · L2525export 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: }
Medium
Install Persistence
Source writes installer persistence such as shell profile or service configuration.
dist/commands/setup-shell.jsView on unpkg · L25dist/worker.jsView file
Findings
3 High6 Medium4 Low
HighTrigger Reachable External Ai Agent Control Surface Mutationdist/worker.js
HighTrigger Reachable Persistencedist/commands/setup-shell.js
HighSemantic Analysis Limiteddist/brain/memvidBrain.js#virtual:normalized:round1
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
LowUrl Strings
Affected versions and remediation
This report applies to agenthub-multiagent-mcp@1.66.0.
See all versions of agenthub-multiagent-mcp for other recorded verdicts.
- Review the evidence and your use of agenthub-multiagent-mcp@1.66.0 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.