optimize-claude-config

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.70). The required workflow embeds and installs an outsider-authored free-text source: .claude/hooks/inject-rules.py reads rule bodies from .claude/rules/**/*.md (which may contain arbitrary text written by non-operating users) and injects that body into the LLM via hookSpecificOutput.additionalContext during PreToolUse(Write).