Proactive associative memory for DSH: system-prompt recall before the model speaks, three-layer auto-consolidation, skill crystallization, and Astra-style context management - handoff ledgers, PLAN whiteboard, water-level sensing. Local-first, model-agnostic, zero deps. 主动联想记忆+Astra 式上下文管理：自动唤回/自动沉淀/技能固化/交接账本与白板跨窗口续命/水位感知。
Use this cross-platform skill in Codex or Claude Code to establish repository-local continuity memory so a future agent can recover objective, status, decisions, validation, risks, and next actions without relying on previous chat history.
DeepSeek Harness plugin: export a session as human-readable Markdown or a self-contained HTML page — full / handoff / readable / audit presets, one-click download
A project-agnostic multi-agent collaboration protocol for the DeepSeek Harness: loop guards that stop runaway agent loops, a six-field handoff format, risk-tiered review routing, and a fixed verify→commit→push→review→close sequence. One skill instead of the same rules copy-pasted into every agent's instructions.
Context compression for DSH: visually configure summary and recent-context budgets (8k + 16k by default), keep tasks on track, and search full conversation history. Friendly to small-context models.
dsh plugin: compiles a rough need into a versioned task book, then hands it to any AI window losslessly via a read-back handshake (/forge /relay /ack /answer /forge-list)
DSH plugin: hands a long session over to a fresh session in the same workspace on /handoff or at a token threshold, carrying the preset, model, permissions and plan mode across; the brief is written by the session's own model and the switch waits for a step boundary instead of interrupting running work.
