skill-issue
Warn
Audited by Gen Agent Trust Hub on Oct 28, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The documentation instructs the user to install a CLI tool from a third-party source using
npm i -g @misha_misha/skill-issueor Homebrew. These sources are outside of the platform's trusted vendor list. - [REMOTE_CODE_EXECUTION]: The skill suggests using
npx @misha_misha/skill-issue, which fetches and executes code directly from the NPM registry at runtime. - [COMMAND_EXECUTION]: The skill provides examples for running commands that access and modify files within sensitive local directories, such as
~/.claude/skillsor~/.codex/skills, which typically house agent logic and configuration. - [INDIRECT_PROMPT_INJECTION]: The tool is designed to process untrusted data (external
SKILL.mdfiles). Because it optionally uses a Large Language Model (--llm) to analyze this content, there is a risk that malicious instructions embedded within a skill being audited could influence the behavior of the auditing process (e.g., bypassing its grading logic or leaking metadata via the LLM API call).
Audit Metadata