changelog-gen

Warn

Audited by Gen Agent Trust Hub on Mar 24, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill uses npx ai-changelog to fetch and run a package from the public NPM registry at runtime. The package is maintained by an unverified third party (LXGIC Studios) and does not originate from a recognized trusted source.- [REMOTE_CODE_EXECUTION]: Executing an unverified package via npx constitutes remote code execution on the user's machine, as the package content is determined by the remote registry and can be updated without the user's knowledge.- [COMMAND_EXECUTION]: The skill relies on shell command execution for both its core functionality (npx ai-changelog) and for retrieving repository history via git log.- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8). Maliciously crafted commit messages in a repository's history could provide instructions that override the AI's intended behavior during the changelog generation process.
  • Ingestion points: Git commit messages retrieved via git log between user-specified refs.
  • Boundary markers: None identified; commit messages are passed directly to the AI for parsing and formatting.
  • Capability inventory: The skill can execute shell commands, access the local file system (git history), and transmit data to external APIs (OpenAI).
  • Sanitization: No evidence of input validation, filtering, or escaping of commit message content is present in the skill instructions.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 24, 2026, 12:24 AM