The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and follow instructions from repository-level files such as AGENTS.md and CLAUDE.md, which creates a surface for indirect prompt injection if those files are malicious.
Ingestion points: Repository files such as AGENTS.md, CLAUDE.md, and host-provided instructions as specified in the SKILL.md workflow.
Boundary markers: No specific delimiters or instructions to disregard embedded commands within these external files are provided.
Capability inventory: The skill possesses the ability to modify project source files and execute local shell commands for validation (e.g., lint, test, and build scripts).
Sanitization: The skill lacks mechanisms to sanitize or validate the instructions contained within these external guidance files before the agent processes them.
[COMMAND_EXECUTION]: The skill instructs the agent to run local development scripts for code validation.
Evidence: SKILL.md and references/stack-guidelines.md direct the agent to execute commands like lint, test, and build as defined in the project's package.json file. This is standard and expected functionality for a frontend development skill.

implement-frontend