The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external content from GitHub and GitLab issues which can be used to influence agent behavior.
Ingestion points: Fetches issue titles, bodies, labels, and comments via gh issue view and glab issue view as described in SKILL.md.
Boundary markers: No explicit delimiters or instructions are provided to the agent to distinguish between valid task data and potentially malicious embedded instructions within the issue content.
Capability inventory: The skill has significant capabilities including code modification, repository cloning, and PR creation using git, gh, and glab CLI tools as documented in SKILL.md.
Sanitization: There are no instructions to sanitize, filter, or validate the fetched issue content before the agent attempts to implement the requested changes.
[COMMAND_EXECUTION]: The skill relies on executing system commands using CLI tools like gh, glab, and git to interact with remote repositories and issue trackers. While these are standard tools for the skill's purpose, they represent the execution path for any instructions derived from untrusted issue data.

implement-issue