The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands like cat, ls, rg, and grep to inspect project documentation (README.md, THEME_GUIDE.md), project configuration (package.json, tsconfig.json), and source code patterns.
[PROMPT_INJECTION]: The skill establishes a surface for indirect prompt injection by reading untrusted project data. Ingestion points: The agent is instructed to read contents from documentation files and project-specific rule files (.cursor/rules/*.mdc). Boundary markers: There are no delimiters or instructions provided to ensure the agent ignores embedded commands or directives within the ingested project files. Capability inventory: The skill enables full-stack implementation capabilities, including filesystem writes and shell command execution, based on the gathered context. Sanitization: No validation or sanitization is performed on the data read from project files before it is processed as context by the agent.

creative-workflow