chore: align atlas-core handoff and normalize files

chore: bump version to 0.1.16

Refactor atlas env init (#146)

feat: KV cache optimization (#144)

Add learning tracking documentation (#141)

chore: bump version to 0.1.15

Fix learning synthesis issues: session_reward context and Gemini structured outputs

fix: Add structured task/step payload support for BYOA adapters (#137)

Refactor: Extract env.py modules for improved maintainability

Fix discovery experience: synthesis limits and module-level instance support

docs: clarify storage requirements for rewards and learning

docs: add training pipeline documentation links

add docs badge and development section to README

add arxiv badge

docs cleanup