dsh-continual-harness¶

⭐ 4 · ✅ active · workflow


Type	workflow	Category	Workflows
Stars	⭐ 4	Status	✅ active
Author	jasen215	Updated	2026-08-20

One-liner¶

DeepSeek Harness plugin for continual self-evolution: persistent memory, periodic review-and-refine, cross-session shared knowledge, and automatic rollback — a plan→validate→apply→rollback loop driven by a model-callable harness_refine tool.

About¶

A DeepSeek Harness (DSH) plugin for self-improving AI agents, providing continual learning through persistent memory, periodic review and refinement, cross-session knowledge sharing, and automatic rollback on failure. It forms a closed loop of plan → validate → apply → rollback. The design is inspired by the open-source prime-agent from Prime Intellect, a self-improving coding harness.

📦 Install¶

dsh plugin --profile <name> add dsh-continual-harness

🚀 Quick Start¶

- insert:
    - id: continual-harness
      name: dsh-continual-harness
      config:
        defaultGlobal: true

📚 Learn more¶

Architecture

src/ domain.ts event declaration merging (SessionEventMap / MessageSourceMap / cordis Events) types.ts HarnessState / RefinementProposal / RefinementResult and other types storage.ts disk read/write of state and history (atomic writes, corruption degradation, local/global merge, jsonl history) refine.ts validation, application, rollback (baseline conflict detection, version increments, growth limi