release

Warn

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: MEDIUMCOMMAND_EXECUTION
Full Analysis
  • Indirect Prompt Injection (MEDIUM): The skill ingests untrusted data from CHANGELOG.md to generate GitHub release notes and possesses the capability to modify source code and push directly to the main branch. This creates a risk where a compromised changelog could influence release metadata or facilitate the publication of malicious code.
  • Command Execution (LOW): The skill executes local commands for linting, testing, and version control (ruff, pytest, git, gh). While legitimate, this represents a significant local command execution surface.
  • Metadata Poisoning (LOW): The skill relies on specific string formats in init.py and CHANGELOG.md. Manipulation of these files could lead to versioning inconsistencies or broken release pipelines.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 16, 2026, 08:59 AM