Cross-session memory with zero-token deterministic recall (bigram-Jaccard fused with FTS5 BM25 via RRF), a tiered approval gate that auto-confirms only reversible user-anchored writes, and reinforcement that strengthens repeated observations. Crystallizes procedures into SKILL.md files that refine in place, curated through an active-stale-archived lifecycle with reversible archive and rollback. v0.4.2 adds a controlled-prune settings page: reversible soft-delete (tombstone, restorable), pinned three-tier protection, read-only heat ordering, and two-stage preview-then-execute with idempotent apply. Anti-bloat convergence for skills and memory, background per-turn review in an isolated call. Requires Node 22.5+.
- License
- MIT
- Added
- 2026-08-24
GitHub info
- License
- MIT
- Primary language
- JavaScript
- Last push
- Aug 23, 2026, 2:04 AM
- Maintainer
- chenzheshushi-commits
- Added
- 2026-08-24
Install
dsh plugin --profile web add "https://github.com/chenzheshushi-commits/dsh-evolve/releases/latest/download/dsh-evolve.tgz"README badge
Add this Markdown to your plugin README to link back to its listing.
[](https://dshget.com/plugins/chenzheshushi-commits/dsh-evolve)Related plugins
MemoryOpenViking#examples/dsh-memory-plugin
★ 34201OpenViking memory and context bundle for DeepSeek Harness: pre-step auto-recall and profile injection, session capture, `viking://` URI guarding, and recall/write memory tools backed by an OpenViking server.
hindsight#coding-agents
★ 21624Hindsight, agent memory that learns: long-term project memory with auto recall and retain, knowledge pages, deep reflection, and per-repo memory banks.
ReMe#typescript
★ 3367Connects DeepSeek Harness to ReMe's local-first, self-evolving personal knowledge base: automatically captures completed main-agent conversations as user-owned Markdown memory, searches conversations and source material through reme_search with BM25, optional embeddings, and wikilink expansion, and schedules daily memory consolidation.