chenzheshushi-commits/dsh-evolve ↗★ 11
dsh-evolve
Self-evolving memory + skill lifecycle for DeepSeek Harness. Cross-session memory with zero-token deterministic recall (bigram-Jaccard fused with FTS5 BM25 via RRF), a tiered approval gate, and reinforcement that strengthens what you repeat. Procedural knowledge crystallizes into SKILL.md files that refine in place and are curated through an active-stale-archived lifecycle (reversible archive, pre-op backups, rollback, never deletes). Includes background per-turn review, anti-bloat convergence for both skills and memory, an auto-grown user profile, and a web settings page. 适合希望Agent具备长期记忆和技能沉淀能力的用户，支持知识自动结晶与强化。
같은 패키지 이름의 다른 저장소
설치
npx -p @deepseek-ai/dsh dsh plugin --profile web add github:chenzheshushi-commits/dsh-evolveConfiguration
Everything is configurable through the plugin's settings page (web profile) or your DSH config. Notable switches:
| Key | Default | Effect |
|---|---|---|
autoConfirmEnabled | true | false = every model write waits for review |
reviewEnabled | true | Background per-turn review |
reviewEveryTurns | 5 | Review throttle |
reviewModel | (main model) | Route review to a different model |
refineLLM | true | Use an LLM pass when crystallizing/refining skills (false = zero-token deterministic assembly) |
reinforceEvery | 3 | Observations per importance step |
memoryMaxChars | 20000 | Memory character budget (0 disables) |
convergeSuggest | true | Surface merge/fold suggestions |
curatorStaleDays / curatorArchiveDays | 30 / 60 | Skill lifecycle thresholds |
ftsEnabled | true | false = pure bigram recall, no SQLite |
The LLM is only ever used for optional auxiliary passes — skill refinement, background review, and skill merging. All of them are single-shot, skippable, and fall back to deterministic behavior on failure. Nothing runs in your main loop.
New configuration
| Key | Default | Effect |
|---|---|---|
refineLLM | true | LLM pass when crystallizing/refining; false is zero-token assembly |
tier1Enabled | true | Always-on preference snapshot at turn start |
maxPendingQueue | 50 | Hard cap on the pending review queue |
disposalMinIdleDays | 30 | Days of zero use before a memory is a cold candidate |
idleMinutes | 5 | Idle delay before suggest/tidy recompute |
approvalPromptEnabled | false | Ask in-turn for direct high-value memory writes |
language | follow-host | follow-host, en, or zh for this page and the injected prompts |
Every key here was already configurable; what changed is where the value is stored and how it reaches the running plugin. Governance knobs nobody has asked to tune at runtime (curator thresholds, heat decay, memory budget, retrieval knobs) remain ordinary YAML config and are not part of the settings form.
New configuration
| Key | Default | Effect |
|---|---|---|
language | follow-host | follow-host, en, or zh. Follows the DSH host locale unless overridden |
Everything configurable in v0.6.x still applies unchanged:
| Key | Default | Effect |
|---|---|---|
disposalMode | manual | manual, suggest, or recoverable tidy |
tidyMaxPerRun | 5 | Maximum automatic soft-deletes per idle run |
idleMinutes | 5 | Idle delay for suggest/tidy |
skillProposalMode | inherit | Follow memory approval mode or explicitly select a skill mode |
skillAutoMaxChars / skillMaxChars | 10000 / 40000 | Automatic/human skill-body limits |
approvalPromptEnabled | false | Ask in-turn for direct high-value memory writes |
approvalPromptMaxPerTurn | 1 | Popup budget per session turn |
New configuration
| Key | Default | Effect |
|---|---|---|
disposalMode | manual | manual, suggest, or recoverable tidy |
tidyMaxPerRun | 5 | Maximum automatic soft-deletes per idle run |
idleMinutes | 5 | Idle delay for suggest/tidy |
skillProposalMode | inherit | Follow memory approval mode or explicitly select a skill mode |
skillAutoMaxChars / skillMaxChars | 10000 / 40000 | Automatic/human skill-body limits |
approvalPromptEnabled | false | Ask in-turn for direct high-value memory writes |
approvalPromptMaxPerTurn | 1 | Popup budget per session turn |