Back to skill

Security audit

Bibverify

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward BibTeX verification helper with disclosed CLI/MCP use and no hidden or destructive behavior in the artifact.

Before installing, verify that `bibverify` is the package you intend to use, consider installing it in a virtual environment, and change the sample config values such as language, BibTeX path, and email to match your project.

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:33
Finding

Unpinned Third-Party Package Installation

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, lines 33–37
Vulnerability Type: Unpinned and automatically upgraded third-party dependency
Risk Level: Medium

Complete Code Snippet:

markdown
For first-time setup, tell the user to install Bibverify from PyPI and create a minimal config:

```bash
pip install -U bibverify
text

### Technical Analysis

The documented installation command retrieves the latest available `bibverify` package from PyPI without pinning a reviewed version or verifying an integrity hash. The `-U` option also directs pip to replace an existing installation with a newer release.

Because the resolved package content can change after this Skill has been reviewed, following the instruction creates a supply-chain trust boundary outside the audited project. A compromised publisher account, compromised package release, or maliciously altered distribution could introduce code that executes during package installation or when the installed CLI is subsequently invoked.

The audit did not establish that the current `bibverify` package is malicious. The finding concerns the mutable, unverified dependency installation mechanism.

### Attack Path

1. An attacker compromises the upstream package publisher, release process, or distribution channel and publishes a malicious `bibverify` release.
2. A user follows the Skill's first-time setup instructions and executes `pip install -U bibverify`.
3. Pip resolves and downloads the attacker-controlled latest release because no trusted version or hash is specified.
4. Package-controlled installation or build behavior executes with the privileges of the invoking user, or the malicious code executes when the `bibverify` command is later used.
5. The payload can access resources available to that user and perform actions within the surrounding Python environment.

### Impact Assessment

Successful exploitation could permit arbitrary code executio
...[truncated 458 chars]
Remediation
View remediation

Remediation Suggestions

  • Pin Bibverify to a specifically reviewed release, for example bibverify==X.Y.Z, rather than installing an unconstrained latest version.
  • Use a requirements or lock file containing cryptographic hashes, and install with pip's --require-hashes option where practical.
  • Verify and document the official package name, publisher, and project URL to reduce typosquatting and package-substitution risk.
  • Remove -U from the default setup command so an existing reviewed version is not silently replaced.
  • Recommend installation in a dedicated virtual environment with ordinary user privileges rather than a global or privileged Python environment.
  • Establish an explicit dependency-update review process before changing the pinned version or accepted hashes.
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)

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

The sample config sets "language": "EN", which imposes a specific language setting in natural-language guidance. The file does not offer an opt-in choice or explain why English is required, so it can violate locale/language policy expectations.

Content

No source excerpt is available for this finding.