Back to skill

Security audit

MIT Reasoning Lite

Security checks for vulnerabilities and agentic risk

Overview

This is a simple reasoning-prompt skill with a disclosed but mutable npx install command and a local installer that appears unreliable rather than harmful.

Before installing, prefer the manual copy path or a pinned, reviewed ClawHub installer version instead of `clawhub@latest`. Treat the included `install.sh` as incomplete because it refers to files not present in the artifact, but its visible commands are limited to creating a user-local directory and copying local files.

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:31
Finding

Execution of an Unpinned, Mutable Package Release

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, line 31
Vulnerability Type: Supply-chain risk caused by unpinned package execution
Risk Level: Medium

Vulnerable Code:

bash
npx clawhub@latest install mit-reasoning-lite

Technical Analysis

The documented installation command instructs users to execute the mutable latest release of the third-party clawhub package through npx. Depending on the local npm configuration and cache, npx can retrieve the package from a remote registry and execute its CLI code with the invoking user's privileges.

Because latest is a mutable distribution tag rather than an immutable version, the code executed in the future may differ from the version originally reviewed. The command also provides no package-integrity value or other verification mechanism. The remotely retrieved package is not included in this project, so its implementation and lifecycle behavior cannot be validated from the audited artifact.

This is an insecure dependency execution pattern, not evidence that the current clawhub package is malicious.

Attack Path

  1. An attacker compromises the package publisher account, registry delivery path, or another component involved in publishing the clawhub package.
  2. The attacker publishes a malicious package release and assigns it to the latest distribution tag.
  3. A user follows the installation command in SKILL.md.
  4. npx downloads or resolves the attacker-controlled release.
  5. The package's CLI or applicable package execution logic runs with the user's permissions.
  6. The malicious release can access or modify resources available to that user.

Impact Assessment

Successful exploitation could allow arbitrary code execution under the account that invokes npx. The reachable scope may include user-owned files, source repositories, environment variables, locally accessible credentials, network resources, and writable configur ...[truncated 288 chars]

Remediation
View remediation

Remediation Suggestions

  1. Replace the mutable latest tag with an exact, reviewed package version:
    bash
    npx clawhub@X.Y.Z install mit-reasoning-lite
    
  2. Verify the selected release's provenance and integrity before recommending it.
  3. Use npm lockfiles and integrity metadata where the installation workflow permits.
  4. Document the expected trusted registry and avoid allowing silent resolution from arbitrary registry mirrors.
  5. Prefer a locally reviewed installation path when feasible.
  6. Correct the local installer's references to absent USAGE.md and README.md files before presenting it as an alternative, because set -e currently causes that installer to terminate partway through execution.
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)

Rp1

Medium
Category
MCP Rug Pull
Confidence
95% confidence
Finding

The installation command uses npx clawhub@latest, which fetches and executes the most recent package version without pinning. This creates a supply-chain risk: if the package is compromised or a breaking/malicious update is published, users of the skill may execute unreviewed code during install.

Content

No source excerpt is available for this finding.