The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because it retrieves and displays content from GitHub issues, which are external and untrusted sources. \n
Ingestion points: Issue content is ingested through gh issue view and web retrieval tools like WebFetch. \n
Boundary markers: No specific delimiters or instructions are present to isolate retrieved content from system instructions. \n
Capability inventory: The skill uses the Bash tool, allowing for significant command execution if the agent is manipulated. \n
Sanitization: No sanitization of the retrieved issue data is described. \n- [EXTERNAL_DOWNLOADS]: The documentation includes standard instructions to download the GitHub CLI and its official keys from trusted sources. \n
Evidence: references/gh-cli-guide.md details how to download the GitHub CLI keyring from cli.github.com and install the package via system managers. \n- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands for issue search and provides installation steps requiring administrative access. \n
Evidence: SKILL.md and references/gh-cli-guide.md demonstrate the use of the Bash tool for API interaction and package installation using sudo.

github-issues