Kb Eval Repo

• Outcome: Review
• Latest audit: Jun 25, 2026 · 9:07 PM UTC
• Version: 1.0.0

Overview

The skill mostly matches its stated repository-evaluation purpose, but it uses an admin Gitea token and writes KB review pages by default.

Install only if you are comfortable giving this skill admin-level access to your Gitea instance and allowing it to create or update review pages in selected KB repositories. Prefer running it with writeReview set to false unless you explicitly want persistence, and use a narrowly scoped bot token instead of an admin token where possible.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands

Findings (11)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 88% confidence
• Finding: The skill documentation declares no explicit permissions while instructing use of capabilities that read files, access the network, and potentially rely on environment-provided credentials/helpers. This creates an implicit trust boundary issue: operators and users may approve or invoke the skill as if it were low-privilege when it can actually reach external repositories and local KB content, increasing the chance of overbroad data access or unintended side effects.

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 97% confidence
• Finding: The skill is presented as an evaluation/query tool, but its workflow includes creating a review file in the selected repository by default, which is a state-changing action. That mismatch is dangerous because a caller expecting read-only analysis may unintentionally authorize repository writes, enabling unwanted content injection, audit noise, or abuse of authenticated repository-modification APIs.

Description-Behavior Mismatch

Medium

• Confidence: 92% confidence
• Finding: The skill includes file-write capability to the knowledge-base repository, which is not necessary for merely evaluating a repository's usefulness. This expands the blast radius from read-only analysis to persistent modification of KB content, enabling unauthorized or misleading report creation if the task input or skill behavior is abused.

Context-Inappropriate Capability

High

• Confidence: 98% confidence
• Finding: The code loads a Gitea admin token for a skill whose stated purpose is repository evaluation/querying, giving it far broader authority than needed. If the skill is compromised or misused, an attacker could modify arbitrary repository contents in the Gitea instance rather than only performing read operations.

Context-Inappropriate Capability

Medium

• Confidence: 94% confidence
• Finding: The skill defaults to creating or updating review documents in the KB repository via writeReview=True, which exceeds the expected scope of a query/evaluation tool. Default-on persistence can be triggered without explicit user awareness and may pollute trusted KB content with machine-generated or attacker-influenced material.

Missing User Warnings

Medium

• Confidence: 93% confidence
• Finding: The skill omits a prominent user-facing warning that it may write an evaluation report into the selected repository unless `writeReview` is false. In this context, the omission is more dangerous because the skill operates on knowledge repositories and could persist generated content into user-selected targets without informed consent, leading to accidental modification of trusted KB state.

Missing User Warnings

Medium

• Confidence: 86% confidence
• Finding: The code writes review files back into the KB repository without any user-facing warning or consent flow in this file. In a knowledge-base context, silent persistence is risky because users may expect an evaluation action to be non-mutating, while the skill actually alters trusted documentation state.

Unpinned Dependencies

Low

• Category: Supply Chain

Content

requests
python-dotenv

• Confidence: 95% confidence
• Finding: requests

Unpinned Dependencies

Low

• Category: Supply Chain

Content

requests
python-dotenv

• Confidence: 92% confidence
• Finding: python-dotenv

Known Vulnerable Dependency: requests — 10 advisory(ies): CVE-2014-1830 (Exposure of Sensitive Information to an Unauthorized Actor in Requests); CVE-2024-47081 (Requests vulnerable to .netrc credentials leak via malicious URLs); CVE-2024-35195 (Requests `Session` object does not verify requests after making first request wi) +7 more

High

• Category: Supply Chain
• Confidence: 90% confidence
• Finding: requests

Known Vulnerable Dependency: python-dotenv — 1 advisory(ies): CVE-2026-28684 (python-dotenv: Symlink following in set_key allows arbitrary file overwrite via )

Low

• Category: Supply Chain
• Confidence: 71% confidence
• Finding: python-dotenv

VirusTotal

59/59 vendors flagged this skill as clean.