openclaw-starter-kit

Multi-threaded conversation memory organized by topic instead of by date. Enables seamless context switching, compaction resilience, and sub-agent tracking — all file-based, no code changes to OpenClaw. Use when you want your agent to maintain separate conversation threads per project/topic, survive context compaction without losing the thread, track sub-agent spawns with model/task/duration, and output a structured conversation-map.json for visualization clients.

Build software projects using structured spec-driven development with the Ralph Wiggum pattern. Use when asked to build, implement, or code a project, feature, or application. Covers discovery (requirements gathering), spec generation, AGENTS.md context generation, and autonomous implementation with fresh-context iterations. Supports any language/framework.