Add output styles prompts

It would be helpful if the prompts for the built-in Proactive, Explanatory and Learning output styles were included here. Also, The "learning mode" prompt seems mislabelled, I believe that's a concept in "Claude for Education" (a service level), not Claude Code.

I extracted them and compared with the repo structure. Claude suggested these changes:

File	Content	Action
system-prompt-output-style-proactive.md	intro + # Proactive Style Active + the numbered The user chose continuous, autonomous execution block, inline	new
system-prompt-output-style-explanatory.md	intro + # Explanatory Style Active + ${INSIGHTS_INSTRUCTIONS}	new
system-prompt-output-style-learning.md	intro + Learn-by-Doing body + ## Insights + ${INSIGHTS_INSTRUCTIONS}	rename of system-prompt-learning-mode.md
system-prompt-output-style-insights-fragment.md	the ## Insights block — interpolated by both Explanatory and Learning	rename of system-prompt-learning-mode-insights.md
system-prompt-output-style-injection-wrapper.md	# Output Style: ${OUTPUT_STYLE_NAME} + ${OUTPUT_STYLE_PROMPT}	new
system-reminder-output-style-active.md	add the one Proactive turn-reminder sentence; no other built-in defines one	amend
system-prompt-auto-mode.md	the numbered block alone — auto mode's reminder body at 2.1.139, the Proactive body at 2.1.220	delete once inlined above

Already exist and correct at 2.1.220 — leave alone:

File	Holds at 2.1.220
system-reminder-auto-mode-clarification-bias.md	## ${AUTO_MODE_HEADING} + bias-against-clarifying-questions + git-safety paragraphs
system-reminder-auto-mode-consent-flow.md	the consent appendix appended when autoModeConsentFlow is set
system-prompt-autonomous-operation-guidelines.md	You are operating autonomously. The user is not watching… — the autonomy_append prompt section
system-prompt-interactive-agent-intro-output-style-conditional.md	the intro line that branches on whether a style is configured