6 Context Compaction Plugins for DeepSeek Harness
The compaction seam: one summarizing backend driven by token pressure and request-error recovery, an optional model-free tool-result pruner, and a human /compact command. There is deliberately no model-facing compact tool.
What DeepSeek Harness defines at this extension point
- Official slot
packages/compaction- Extension kind
- seam
- ctx keys (2)
ctx.compactionctx.toolResultPruner- Official built-ins
- DeepSeek ships 1 implementation herecompaction-basic, compaction-tool-result-pruner
- Gap
- largehow much of this seam the official taxonomy grades as still open
- Audience
- developer
- Artifact kind
- cordis-plugin
- Seams
- 1
Every field above is read from the DeepSeek Harness (DSH) source tree — deepseek-ai/deepseek-harness — not from our own guesswork about what these projects do. Same data as a live filter: browse this category in the full plugin list.
6 Context Compaction Plugins indexed here
Graph Memory
adoresever
Openclaw记忆插件Knowledge Graph + Memory；Knowledge Graph Context Engine for OpenClaw — extracts structured triples from conversations, compresses context 75%, enables cross-session experience reuse
Mnemon
mnemon-dev
LLM-supervised persistent memory for AI agents — graph-based recall, cross-session knowledge, single binary. Works with DeepSeek Harness, Claude Code, OpenClaw, and any agent runtime.
Dsh Compaction Instant
KitDoesIt
LLM-free lossless* compaction engine for DeepSeek Harness
Dsh Ui Preset Enhance
lssyd20070106
Third-party DSH WebUI enhancement plugin: custom backgrounds, theme colors, prompt presets, token/context visibility, and manual compact.
Dsh Composer History
PerryLink
Terminal-style input history for the DeepSeek Harness web composer: edge-first arrows with exact draft/caret restore, browser-local persisted history, Ctrl+R reverse search, workspace recall - and sliding-context awareness (compaction summaries in recall/search, compaction notice with one-click /compact fill).
Dsh Asc
lmst2
No description provided.