Back to skill

Security audit

Auxiliar Solve

Security checks for vulnerabilities and agentic risk

Overview

The skill has a coherent tool-recommendation purpose, but it asks users to install an unpinned third-party MCP service and lacks clear privacy and consent boundaries for sensitive document workflows.

Review carefully before installing. Pin and verify the npm package and source, run the MCP with least privilege, avoid sending sensitive financial or personal document details unless you accept the third-party handling, and treat any returned install commands as recommendations that need separate validation.

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 (1)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:29
Finding

Unpinned Third-Party npm Package Execution

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, lines 29–33
Vulnerability Type: Supply-chain risk from an unpinned executable dependency
Risk Level: Medium

Vulnerable Code

markdown
### Step 1. Install the auxiliar MCP server

```bash
claude mcp add auxiliar -- npx auxiliar-mcp
text

### Technical Analysis

The documented command configures the agent to execute `auxiliar-mcp` through `npx` without specifying an exact package version or verifying package integrity. Depending on the local npm configuration and cache state, `npx` may retrieve the current package release from the configured npm registry and execute it with the invoking user's privileges.

Because the package implementation is not included in the audited project, its behavior cannot be verified from this repository. The effective executable payload may also change after this skill has been reviewed if a new upstream release is published. A compromised package publisher, npm account, registry, or newly published malicious release could therefore turn the documented setup operation into arbitrary local code execution.

The MCP is also described as returning copy-paste installation commands for other tools. Those secondary commands represent an additional trust boundary and should not be executed automatically without package, source, version, and integrity validation.

### Attack Path

1. An attacker compromises the `auxiliar-mcp` publisher account, package distribution channel, or upstream release process.
2. The attacker publishes a malicious release under the same npm package name.
3. A user or agent follows the skill's documented command without an exact version constraint.
4. `npx` resolves and downloads the mutable package release from the configured registry.
5. The package runs malicious initialization or server code under the account that invoked the command.
6. The payload can access resources available to that account and may recommend additional attacker-controll
...[truncated 697 chars]
Remediation
View remediation

Remediation Suggestions

  1. Replace the unversioned package reference with an exact, reviewed version, for example:

    bash
    claude mcp add auxiliar -- npx --yes auxiliar-mcp@0.1.0
    

    The actual version must be selected only after reviewing that release. Do not use ranges, tags such as latest, or floating versions.

  2. Prefer a locked installation workflow using a committed lockfile. Verify npm integrity metadata and retain the expected package digest in release documentation or a trusted manifest.

  3. Document the authoritative npm package page, source repository, publisher identity, and release-signing or provenance information so users can validate that they are installing the intended package.

  4. Review the package tarball and transitive dependency tree before approving a release. Re-audit whenever the pinned version or dependency lockfile changes.

  5. Run the MCP with least privilege in an isolated environment or container. Restrict filesystem access, environment variables, credentials, and outbound network access to what is strictly required.

  6. Treat installation commands returned by solve_task or related MCP methods as untrusted recommendations. Require explicit user confirmation and validate each package name, source, exact version, integrity value, permissions, and install scripts before execution.

  7. Disable or scrutinize npm lifecycle scripts where operationally possible, and monitor the package for ownership changes, unexpected releases, and provenance failures.

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill encourages installing and querying a third-party MCP service but does not clearly warn that prompts, task descriptions, document metadata, or other sensitive workflow information may be sent to or processed by external infrastructure. For a skill focused on invoices, receipts, bookkeeping, and document extraction, that omission is more dangerous because those workflows commonly involve financial and personal data.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The invocation guidance is broad enough that an agent may call this skill in many generic 'capability gap' or tool-selection situations, increasing the chance it is used automatically and unnecessarily. In context, that matters because using this skill leads to external tool installation and service queries, so overly permissive triggers can expand exposure to third-party code and data sharing.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
95% confidence
Finding

The skill instructs users to install and run an external MCP server via npx auxiliar-mcp without pinning an exact version or integrity hash. That creates a supply-chain risk: future package changes, account compromise, or typosquat/package takeover could cause arbitrary code to run in the agent environment at install or execution time.

Content

No source excerpt is available for this finding.