Security audit

Target Novelty Scorer

Security checks for vulnerabilities and agentic risk

Overview

This skill looks like a demo that presents synthetic biomedical scoring as real literature-mining analysis, so users should review it carefully before relying on it.

Treat this skill as a mock or prototype unless it is updated to call real literature databases and clearly label data provenance. Do not use its scores, paper counts, trial counts, or confidence values for scientific, investment, or business decisions without independent verification and real citations.

SkillSpector

Vulnerability Patterns

Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (5)

Description-Behavior Mismatch

High

Confidence: 98% confidence
Finding: The tool advertises literature-mining novelty scoring, but the core search function generates simulated random values instead of retrieving real PubMed data. This can mislead users into making scientific or business decisions based on fabricated evidence, which is especially risky in a biomedical context where output may be treated as decision support.

Intent-Code Divergence

Medium

Confidence: 94% confidence
Finding: The file-level description and CLI usage present the skill as a real target novelty scorer, while the internal searcher admits it is only a simulated implementation. This inconsistency increases the chance that operators will trust the output as factual, creating a deception/integrity risk even without classic code-execution behavior.

Unpinned Dependencies

Low

Category: Supply Chain
Content: dataclasses numpy
Confidence: 94% confidence
Finding: The dependency is unpinned, so installs may resolve to different versions over time, reducing reproducibility and potentially pulling in a newly vulnerable or incompatible release. While this is a supply-chain hygiene weakness rather than an immediately exploitable flaw by itself, it increases security uncertainty in any environment that installs the skill.

Unpinned Dependencies

Low

Category: Supply Chain
Content: dataclasses numpy
Confidence: 98% confidence
Finding: The numpy dependency is unpinned, which allows future installs to pull arbitrary newer or older versions depending on resolver behavior and package index state. In this file, that risk is elevated because numpy also has known advisories across some versions, so leaving it unpinned makes it harder to ensure deployment avoids vulnerable releases.

Known Vulnerable Dependency: numpy — 10 advisory(ies): CVE-2014-1859 (Numpy arbitrary file write via symlink attack); CVE-2021-41495 (NumPy NULL Pointer Dereference); CVE-2021-33430 (NumPy Buffer Overflow (Disputed)) +7 more

Critical

Category: Supply Chain
Confidence: 91% confidence
Finding: Referencing numpy without a safe version constraint is dangerous here because static analysis identified multiple known vulnerabilities affecting some numpy releases. Since the requirement is completely unconstrained, an environment could install an affected version, exposing the skill to issues such as memory-safety bugs or unsafe file-handling behavior depending on how numpy is used and the runtime context.

Static analysis

No suspicious patterns detected.