madage/dsh-self-improved
Long-term memory & self-evolving plugin for DSH: L0 capture → L1 memory extraction → L2 scene grouping → L3 user persona, auto recall injection, skill synthesis, fully local (SQLite FTS5 + jieba, optional vector recall).
$ dsh plugin --profile web add dsh-self-improved$ dsh plugin --profile web add github:madage/dsh-self-improvedHealth breakdown
80 / 100This plugin has not fully disclosed its data-collection behavior. Review the source and manifest before installing.
Security
Key metrics
Related
Related plugins
Cross-agent, local-first persistent memory plugin for DeepSeek Harness (DSH), powered by Mnemon. It shares long-term memory across Mnemon-enabled agents and adds runtime memory, searchable project documents, semantic recall, knowledge graph, and a Sidebar UI.
Cross-session memory: SQLite with a human-editable Markdown mirror, background consolidation (dedup, merge, conflict resolution), and six memory tools.
Cited memory over DSH's lossless session log: distilled facts carry `(sessionId, eventRange)` citations that expand back to the exact original log excerpt.
ShiGuang Memory Engine (SGME) bridge: multi-agent shared long-term memory via HTTP — L0/L1/L1.5/L2 distillation, scenario-based injection, unified search, and proactive care signals (memory_search / wiki_search / signal_pull / signal_claim / signal_ack), installable as `dsh-sgme`.
Action-state time memory: record typed states and actions, then analyze trends, anomalies, and causality.
Automatic conversation distillation: background subagent reflection + skill create/update.