Back to skill

Security audit

Planet Express Marketplace

Security checks for vulnerabilities and agentic risk

Overview

The skill matches its file marketplace purpose, but it needs review because it guides agents through real blockchain payments and file exchange without explicit consent or content-safety guardrails.

Install only if you intend to let an agent interact with this marketplace. Require explicit approval before any purchase, listing fee, or transaction submission; verify network, price, recipient, and listing details; treat all marketplace files as untrusted; and pin or isolate any SDK installation.

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:100
Finding
Unpinned Third-Party SDK Installation Creates a Supply-Chain Risk## Vulnerability Details **File Location**: `SKILL.md`, line 100 **Vulnerability Type**: Unpinned third-party dependencies from external package registries **Risk Level**: Medium **Vulnerable Code Snippet**: ```text Install SDK: `npm i dropclaw` or `pip install dropclaw` ``` ### Technical Analysis The documentation instructs users to install the `dropclaw` package from either npm or PyPI without specifying an audited version, integrity hash, lockfile, registry URL, or publisher verification procedure. These commands therefore resolve the package dynamically at installation time. Because the installed artifact can change after this Skill has been reviewed, package-owner compromise, registry-account takeover, or publication of a malicious future release could cause users to retrieve attacker-controlled code. Depending on the package manager and package contents, malicious code could run through npm lifecycle scripts during installation or when the installed SDK is imported or invoked. The project does not itself contain or execute a malicious payload. The issue is the unsafe dependency-installation recommendation and its exposure to an external software supply chain. ### Attack Path 1. An attacker compromises the relevant npm or PyPI publisher account, registry artifact, or upstream release process. 2. The attacker publishes a malicious version of the package under the expected `dropclaw` name. 3. A user follows the documented `npm i dropclaw` or `pip install dropclaw` instruction. 4. Since no version or integrity value is pinned, the package manager resolves and downloads the attacker-controlled release. 5. Malicious npm installation hooks may execute during installation, or malicious package code may execute when the SDK is imported or used. 6. The payload runs with the privileges and environmental access of the installing user or subsequent SDK process. ### Impact Assessment Successful exploitation could permit a ...[truncated 437 chars]
Remediation
## Remediation Suggestions 1. Replace floating installation commands with exact, previously audited package versions, such as `dropclaw@X.Y.Z` for npm and `dropclaw==X.Y.Z` for Python. 2. Provide lockfiles and verify package integrity using trusted checksums, npm integrity metadata, or hash-checked Python requirements. 3. Document the expected package publisher, official registry location, source repository, and release-signing or provenance information. 4. Review each dependency update before changing the pinned version, including package ownership, lifecycle scripts, transitive dependencies, and release provenance. 5. Recommend installation in an isolated, least-privileged environment without wallet secrets or unrelated credentials. 6. Disable unnecessary npm lifecycle scripts during installation where compatible, and use reproducible installation mechanisms such as `npm ci` or `pip install --require-hashes`.
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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
Confidence
95% confidence
Finding
The skill directly instructs users to initiate paid actions and exchange file content, but it does not prominently warn that these operations can incur real financial charges or that purchased/uploaded files may contain untrusted or harmful content. In an agent setting, insufficient disclosure increases the risk of unintended payments, unsafe handling of external files, and user confusion about when money will be spent.

Static analysis

No suspicious patterns detected.