Security audit

Vocab Voyage Mcp

Security checks for vulnerabilities and agentic risk

Overview

The skill appears to be a legitimate vocabulary MCP service, but its setup and tracking details need review before installation.

Review the privacy and installation implications before using it. Prefer the direct hosted MCP connector or OpenClaw install path over the unpinned npx bridge, avoid putting bearer tokens in command-line arguments when possible, and treat the service as recording usage/source metadata on tool calls.

A.I.G

Vulnerability Patterns

Insecure DependenciesIntroduces malicious components through unsafe dependency sources
Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally

Findings (2)

T08 · Insecure Dependencies

Warning

Location: README.md:66
Finding: Unpinned Third-Party Package Is Automatically Downloaded and Executed
Content: View full analysis
Remediation: View remediation

other

Note

Location: SUBMISSION_LINKS.md:1
Finding: Per-Call Usage and Client Attribution Telemetry Is Not Disclosed in the Primary Skill Description
Content: View full analysis
` query param. The MCP server records this on every tool call into `mcp_tool_calls.install_source`, so you can see which channel actually drives usage. The server normalizes refs to lowercase `[a-z0-9_-]{1,32}`. The slugs below are the canonical ones — match these exactly so the Friday analytics query buckets cleanly. ``` The documented analytics query shows additional per-call data used for aggregation: ```sql select coalesce(install_source, '(direct/no-ref)') as source, count(*) as calls, count(distinct user_id) as authed_users, count(distinct ip_hash) filter (where user_id is null) as anon_clients, round(avg(latency_ms))::int as avg_latency_ms, round(100.0 * count(*) filter (where status = 'ok') / count(*), 1) as success_pct from public.mcp_tool_calls where created_at >= now() - interval '7 days' group by 1 order by calls desc; ``` ### Technical Analysis The repository states that the hosted server records the installation source on every tool invocation. The analytics query further demonstrates that stored tool-call records include or support association with: - An authenticated `user_id`. - A hashed IP identifier for anonymous clients. - Installation or referral source. - Request latency. - Request status. - A creation timestamp used for time-window analysis. Hashing an IP address does not necessarily make the resulting identifier anonymous. A stable hash can still permit repeated calls from the same client to be correlated, and IP addresses have a relatively constrained input space when no secret-keyed construction i ...[truncated 1870 chars]
Remediation: View remediation

SkillSpector

Vulnerability Patterns

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (1)

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 92% confidence
Finding: The document explicitly instructs third parties to include attribution identifiers via ?ref= or x-install-source, and states these are recorded on every tool call for analytics. This creates undisclosed telemetry linkage at the point of installation/use, which can undermine user privacy expectations and enable tracking of acquisition source without clear end-user notice or consent.
Content: No source excerpt is available for this finding.