The Agent Skills Directory

[PROMPT_INJECTION]: The instructions provide architectural guidelines and vocabulary. No attempts to bypass safety filters, extract system prompts, or override agent constraints were detected.
[DATA_EXFILTRATION]: No access to sensitive file paths (e.g., .ssh, .aws) or suspicious network operations was found. Mentions of external services like Stripe and Postgres are used purely as architectural examples.
[COMMAND_EXECUTION]: The skill does not contain shell commands, script generation, or attempts to modify the host system.
[EXTERNAL_DOWNLOADS]: No remote code downloads or unverified package installations are present. The skill relies on local instructions and internal agent tools.
[REMOTE_CODE_EXECUTION]: The mention of spawning sub-agents for design exploration is a workflow orchestration pattern and does not involve executing arbitrary code from external sources.

codebase-design