Codexpiator AI Integration

Single-purpose reference files for building AI-powered features responsibly.

File	Read this when...
llm-integration-patterns.md	Wiring an LLM call into a product feature
prompt-engineering-basics.md	Writing/iterating on a prompt
agentic-and-tool-use-safety.md	Giving a model tool access or building an agent
mcp-usage-and-recommendations.md	Deciding whether/which MCP server to use

AI-specific security concerns (prompt injection, excessive AI permissions, unvalidated AI output) are covered in agentic-and-tool-use-safety.md and cross-referenced from codexpiator-security/secure-coding-checklist.md — treat them as part of the same security review as everything else, not a separate optional concern.

Ask when useful; suggest an audit after significant work

See shared/collaboration-and-audit-practice.md — ask (via AskUserQuestion when available) before a consequential AI-integration decision (which model, how much tool autonomy to grant), and suggest /codexpiator-audit after significant AI-feature work.