The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to process and implement work described in user-provided specifications or tickets, which creates a surface for indirect prompt injection.
Ingestion points: The instructions in SKILL.md explicitly direct the agent to "Implement the work described by the user in the spec or tickets."
Capability inventory: The agent is instructed to execute shell commands (zsh), run tests, perform typechecking, use formatters, and commit work to a git repository.
Boundary markers: The instructions lack explicit requirements for using delimiters or "ignore embedded instructions" warnings when processing the untrusted input.
Sanitization: There are no defined steps for sanitizing, escaping, or validating the content of the specs or tickets before the agent acts upon them.
[METADATA_POISONING]: The LICENSE.mattpocock file includes a copyright date of 2026. This future-dated metadata is a minor inconsistency that could be deceptive, though it is likely a simple typographical error.

implement