Our world in data

• Outcome: Pass
• Latest audit: Sep 10, 2026 · 8:09 PM UTC
• Version: 0.1.3

Overview

This skill is a straightforward OWID lookup helper with disclosed use of a third-party Python package and no hidden persistence or data-harvesting behavior in the artifact.

Install the dependency only if you are comfortable trusting owid-catalog and its dependencies, preferably in a virtual environment. Expect queries to be performed in English and ask the agent to translate the final answer when needed.

A.I.G

Vulnerability Patterns

• Insecure Dependencies: Introduces malicious components through unsafe dependency sources
• Skill Instruction Hijacking: Alters the agent's session goals or safety constraints when the skill loads
• Agent Memory Poisoning: Writes attacker-controlled rules into memory that affect later sessions
• Remote Payload Retrieval and Execution: Fetches external code whose behavior can change after review
• Embedded Malicious Code: Ships malicious scripts inside the skill and executes them locally

Findings (1)

T08 · Insecure Dependencies

Warning

• Location: SKILL.MD:27

Finding

Unverified Third-Party Release Candidate Installation

Content

View full analysis

Vulnerability Details

File Location: SKILL.MD, lines 27–30
Vulnerability Type: Third-party software supply-chain risk
Risk Level: Medium

Vulnerable Code

markdown

## Installation

```bash
pip install owid-catalog==1.0.0rc2

text

### Technical Analysis

The skill directs users or agents to install and subsequently import a third-party Python package. Although the package version is pinned, it is a release candidate, and the installation command does not verify artifact hashes. The project also supplies no lockfile, transitive-dependency constraints, trusted-index configuration, provenance validation, or isolated-environment requirement.

Version pinning limits direct version drift but does not verify the integrity of the selected artifact or its transitive dependencies. Python package installation may execute package build hooks, while later importing the installed package executes its initialization code. Therefore, compromise of the specified distribution, one of its dependencies, or the configured package index could result in local code execution.

The audit did not establish that `owid-catalog` is malicious. This finding concerns the unsafe trust and installation model documented by the skill.

### Attack Path

1. An attacker compromises the selected package artifact, one of its transitive dependencies, or the package source used by `pip`.
2. A user or agent follows the skill's installation instruction.
3. `pip` downloads the affected component without validating project-supplied cryptographic hashes.
4. Malicious code executes during package build or installation, or when the documented `owid.catalog` module is imported.
5. The code runs with the permissions and environmental access of the invoking user or agent.

### Impact Assessment

Successful exploitation could provide arbitrary code execution under the invoking account. Depending on that account's permissions and environment, an attacker could read or alter accessible fi
...[truncated 298 chars]

Remediation

View remediation

Remediation Suggestions

1. Use a reviewed stable package release rather than a release candidate where practical.
2. Generate a dependency lockfile that constrains all transitive dependencies.
3. Record cryptographic hashes and install with hash enforcement, such as pip install --require-hashes -r requirements.txt.
4. Require installation inside a dedicated, least-privileged virtual environment or container.
5. Explicitly configure and document the trusted package index instead of relying on ambient pip configuration.
6. Validate package provenance and review package metadata, source, and dependency changes before upgrading.
7. Do not permit an agent to install the dependency automatically without explicit user approval.
8. In higher-assurance environments, build and scan wheels internally and install only from an approved artifact repository.

SkillSpector

Vulnerability Patterns

• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (1)

Natural-Language Policy Violations

Medium

• Category: Not specified by scanner
• Confidence: 95% confidence

Finding

The language policy requires all searches to be performed in English even when the user asks in another language. This is a natural-language locale policy constraint presented as mandatory behavior, and the file does not offer a user choice or opt-in for that restriction.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.