two-vendors-one-package-kmp-api-skew

Pass

Audited by Gen Agent Trust Hub on Aug 30, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill contains multiple shell-based auditing commands in the 'Verifying it' section designed to be run from a repository root.
  • Evidence: The scripts utilize grep, sed, awk, and sort to parse project configuration files and source code.
  • Context: These operations are limited to local file analysis and are standard for software engineering auditing tools.
  • [INDIRECT_PROMPT_INJECTION]: The provided scripts define a surface for ingesting data from project files, creating a potential vector for indirect injection if project files contain malicious content designed to exploit text-processing utilities.
  • Ingestion points: gradle/libs.versions.toml, build.gradle.kts, and all Kotlin source files (*.kt) within the project tree.
  • Boundary markers: The analysis relies on regular expression matching for import statements and artifact coordinates, and checks for preceding comments (//) as validation markers.
  • Capability inventory: The scripts use grep, sed, and awk to extract, transform, and report on file contents.
  • Sanitization: No sanitization or escaping is applied to the ingested file content before it is processed by the shell utilities.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 30, 2026, 07:04 AM
Security Audit — agent-trust-hub — two-vendors-one-package-kmp-api-skew