The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructions require the agent to execute shell commands to run and verify tests during the development cycle.
Evidence: Found in SKILL.md ("Run the narrowest relevant test command", "Report the... exact commands run").
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted external data (such as tickets, specifications, and architecture documents) and possesses command execution capabilities, creating a surface for indirect prompt injection.
Ingestion points: SKILL.md refers to reading authoritative tickets, specs, repository instructions, and domain/architecture documents.
Boundary markers: The skill includes logic to reject sources marked with kind: "spec-explainer" and mandates resolving authoritative specs via derived_from fields.
Capability inventory: Shell command execution for running test suites.
Sanitization: No explicit sanitization or escaping of external content before processing is mentioned, although it emphasizes limiting execution to "narrowest relevant test commands."

tdd