The Agent Skills Directory

[COMMAND_EXECUTION]: The skill performs standard software development operations, including running test suites, typecheckers, linters, and Git commands for branch and worktree management. These actions are within the expected scope for a skill designed to implement technical specifications (SKILL.md, worktrees.md).\n- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data, such as tickets and specifications, which could potentially contain malicious instructions.\n- Ingestion points: Technical specifications, issue tracker tickets, and repository documentation (SKILL.md).\n- Boundary markers: Absent. No specific boundary markers or delimiters for untrusted content are defined in the instructions.\n- Capability inventory: The skill possesses capabilities for file system modification, Git operations (commits, PR creation), and arbitrary shell command execution for testing and cleanup (SKILL.md, worktrees.md).\n- Sanitization: Absent. No explicit sanitization or filtering of ticket content is specified before the agent acts upon it.

implement-spec