MimicHunterZ/dsh-agent-compact

README excerpt

agent-compact English · 简体中文 Context compression for DeepSeek Harness: lets the agent autonomously call context compact to compress a span of the conversation it chooses — the finished, no-longer-needed middle — and replace it with a checkpoint the agent writes itself. Why Compaction is normally a full-context sweep: the official engine only ever compresses from the start of the conversation, so the opening's task plan and direction are partially lost along with the compressed information . context compact compresses only the span the agent selects — a finished step, a debugged log exchange, an off-track discussion — while the important opening and the recent context stay intact. Span compaction keeps information loss from compaction as small as possible — like human memory, the middle is not compressed indiscriminately: what can be summarized is consolidated into a checkpoint, and important details stay word-for-word — the agent decides what is truly dead, and only that gets condensed. Typical moments to use it: - a task step is done — compress it, keep the remaining steps and the active instruction live; - a bug hunt or a wrong research direction is over — compress that exchange …

Category