The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill executes arbitrary commands specified in ticket acceptance criteria and configuration files such as test_command and push_check. This allows data from the issue tracker or project configuration to dictate shell execution on the host system.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from ticket bodies, pull request comments, and pull request text while possessing high-privilege capabilities including shell execution, git operations, and GitHub API access. * Ingestion points: Ticket bodies (What to build, Acceptance criteria), PR comments, PR bodies, and docs/dev-agents/config.md. * Boundary markers: The instructions include a specific directive to treat external content as data rather than instructions, providing a prompt-level safeguard. * Capability inventory: Shell command execution (via tests and checks), git operations, gh CLI management, and wkc-tracker operations. * Sanitization: No explicit sanitization or validation of the commands extracted from tickets is described.
[COMMAND_EXECUTION]: The skill makes extensive use of shell commands through git, gh, and user-defined test/check commands, which are the primary vectors for the identified dynamic execution risks.

wkc-implement