The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection by design, as it fetches and processes external data to drive its implementation logic.
Ingestion points: The skill fetches ticket details from an issue tracker based on user references and reads specification files (SKILL.md, TRADUCAO.pt-BR.md).
Boundary markers: There are no explicit instructions or delimiters provided to the agent to treat external ticket content as untrusted data or to ignore embedded instructions within those tickets.
Capability inventory: The skill writes code to the local filesystem (Step 1), executes tests and typecheckers via the /tdd skill (Step 1), performs refactoring logic (Step 3), and modifies the repository state using git commit (Step 5).
Sanitization: The skill lacks any mentioned sanitization, filtering, or validation steps for content retrieved from external issue trackers.
[COMMAND_EXECUTION]: The skill uses shell-level Git commands to manage snapshots and finality during the development process.
Evidence: Usage of git stash create and git restore --source=<snapshot> --worktree -- <files> for refactoring safety (Step 3), and git commit for final reporting (Step 5).

implement