The Agent Skills Directory

[Indirect Prompt Injection] (LOW): The skill retrieves data from GitHub issues, which are external and potentially attacker-controlled sources.
Ingestion points: The gh issue list command in SKILL.md is used to fetch the body, title, and labels fields from issues.
Boundary markers: Absent. The retrieved content is passed to the agent context without delimiters or protective framing.
Capability inventory: This skill is restricted to read-only operations and data parsing; it does not contain capabilities for writing files or executing commands based on the fetched data.
Sanitization: Absent. Raw text from GitHub is processed and presented to the agent without filtering.
[Unverifiable Dependencies & Remote Code Execution] (INFO): The skill relies on external command-line tools gh and jq.
Evidence: The documentation provides clear, manual installation instructions for users via trusted system package managers (Homebrew, apt, winget), which is considered safe practice and does not involve automated remote script execution.

github-issue-fetcher