The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes local shell commands such as ls, cat, and grep to inspect the current project directory, detect the structure, and read existing configuration files.
[COMMAND_EXECUTION]: As part of its documented cleanup process, the skill executes file and directory deletions for legacy configuration files, specifically rm -rf .cursor/rules/ and rm .cursorrules.
[PROMPT_INJECTION]: The skill identifies an indirect prompt injection surface by ingesting content from untrusted local files.
Ingestion points: Reads existing CLAUDE.md and AGENTS.md files using cat in SKILL.md (Step 1).
Boundary markers: None are present; the skill reads file contents directly into the agent's context without delimiters or instructions to disregard embedded commands.
Capability inventory: The skill is capable of modifying and creating files (CLAUDE.md, AGENTS.md) and deleting directories (rm -rf).
Sanitization: No sanitization or validation of the content from the ingested files is performed before it is used to determine the setup logic.

peach-setup-ui-proto