101 results
DeepSeek Harness host plugin: archive raw tool results into the session folder, clear them from the visible conversation at round boundaries (or per-step in overclock mode), and let the model recall them through a read_tool_result_log(turn, step) tool.
Manual budget-targeted context compression for DeepSeek Harness: the conversation model picks ranges, a cheap flash-tier route writes the checkpoint summaries.
Tiered micro-clear layer for DeepSeek Harness compaction: clears whitelisted re-runnable tool results to placeholders before the official summarization tier, freeing context with no model call.
Compact, searchable, tag-based agent preset management for DeepSeek Harness
Active Context Pruning (ACP) for the DeepSeek Harness — model-driven context management as a CompactionEngine backend.
AgentFrame compaction backend for DeepSeek Harness: semantic + physical KV compression (28.4x) replacing LLM summarization
Browser-style UI zoom for DeepSeek Harness and DSH Desktop — Ctrl +/−/0, Ctrl+wheel, and a compact zoom widget with a saved level.
Replace the DSH Web reasoning-effort chooser with a compact animated FX slider that scales with effort level.
Adaptive task and risk routing, loop budgets, and progressive tool-result pruning for DeepSeek Harness
High-fidelity, faithful, bilingual, recursive compaction backend for DeepSeek Harness — a drop-in upgrade over dsh-compaction-basic.
A compact DSH model and reasoning quick-switcher showing the last three choices.
Document-review modal for DeepSeek Harness web: design documents and plan reviews open in a full rendered-markdown modal instead of the compact composer card.
Model-scoped DSH plugin: GPT Responses lifecycle, Native V2 compaction and Hosted/Alpha Search; Grok native Web/X Search
提问面板折叠：agent 提问时把问答面板折叠成一条紧凑条（保留标题与取消），展开后回答草稿不丢。Collapse the DSH question panel into a compact strip; the unsent answer draft survives expand.
Idle-triggered context compaction for DeepSeek Harness: compact a session once it has gone quiet past a token floor
Keeps a local model's context at a size the GPU handles well: measured prefill speed, hard ceiling, early compaction for DeepSeek Harness
Session health for DeepSeek Harness: real token-meter data, configurable continue-vs-new decision model. Reactive header badge (compaction-aware), /compass command, context_compass tool.
Workspace History: journals every compaction summary to <workspace>/.dsh/history/<unix-ts>.<short-session-id>.md, and adds a History subtab to the Workspace Overview tab for reading the journal. Per-session toggle lives in the granular Settings tab — requires dsh-granular-settings (hard dependency).
Chat branch tree plugin for DeepSeek Harness: mind-map of conversation branches, group folders, archiving and compaction preview.
Durable DeepSeek Harness bundle: the local-model performance tuning carried as a re-installable plugin layer (compaction, tool-result pruning, time context, cloud title routing, text-toolcall guard) — survives dsh updates
自动上下文策展：tool 输出瘦身 + 无效日志清理 + 工作流/记忆钉扎保护 + 先归档后压缩
DSH web plugin: shrink chat content density — markdown typography tokens, message-body spacing and the user bubble, live-adjustable from a card under Settings → Plugins. Host half registers the settings namespace that makes the Plugins tab dispatch the card.
Composer control for DeepSeek Harness auto-compaction: set a percent (30–80) or absolute token threshold that persists in settings.yaml.
DeepSeek Harness plugin: aggressive token optimization without affecting task quality. Trims tool results, manages context window pressure, optional LLM-summary compaction, stable prompt-cache layout.