The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface. It retrieves and processes issue descriptions and comments using the gh CLI and treats this untrusted external content as requirements for code implementation.
Ingestion points: The body and comments fields from the gh issue view command output in SKILL.md.
Boundary markers: None. The skill does not implement delimiters or explicit instructions to ignore potentially malicious directions embedded within the issue content.
Capability inventory: The skill possesses extensive capabilities, including filesystem modification (Edit, Write tools) and shell command execution (Bash tool), which are used to implement and verify changes.
Sanitization: There is no evidence of input validation or sanitization for the data fetched from GitHub before it is interpreted by the agent.
[COMMAND_EXECUTION]: The skill dynamically identifies and executes various system commands for project verification (e.g., npm test, cargo check, go test, pytest) based on the project's file structure. While these are standard development tools, they are executed in a workflow directly influenced by untrusted external issue data.

github-do-issue