dsh-akn-plugin
Local-first Agent Experience Network for DeepSeek Harness, distilling task-level experiences from traces and live manifests, grading by H0-H4 evidence, discovering by compatibility, and promoting only after explicit review.
DeepSeek Harness 的本地优先 Agent 经验网络：从 trace 与实时 Manifest 蒸馏任务级经验，按 H0-H4 证据分级、按兼容性发现，经显式审阅后才晋升。
How to install
dsh plugin add dsh-akn-plugin About
Agent Experience Network This repository is the public dsh-akn-plugin distribution and the AEN/AEXP reference implementation. It is maintained jointly by symmetryseeker and Jiaoyang Li (@jiaoyangli-shadow7day). See Maintainers and the migration record. Agent Experience Network（AEN）是一个处于 Draft/Pilot 阶段的开放协议与实现，目标是让不同开发者的 Agent 在明确的 Model × Harness × Environment 适用范围内，共享经过审查和验证的任务级执行经验。 项目不会把每次工具调用发布成知识，也不会默认上传原始 Trace。Trace 是证据来源之一；可分享的 Experience 必须同时描述任务、Model、Harness Manifest、结果、正反例、适用边界和证据强度。 30 秒理解这个项目 当一个 Agent 完成任务、从失败中恢复或发现某种配置不适用时，真正值得共享的不是全部日志，而是少量经过选择和审阅的任务经验：它在什么任务、Model、Harness 和环境…
Recommendation signals
Meta
- License
- Apache-2.0
- Language
- TypeScript
- GitHub stars
- 1
- mo. downloads
- 201
- Last push
- 2026-08-20
- Created
- 2026-08-18
Basic safety check
- Findings
- None
- Sources
- curated:awesome-dsh-plugin.com, curated:awesome-dsh-plugin/awesome-dsh-plugin
- Topics
- dsh-plugin
Related plugins
hindsight
vectorize-io/hindsight/tree/main/hindsight-integrations/coding-agents
Hindsight, agent memory that learns: long-term project memory with auto recall and retain, knowledge pages, deep reflection, and per-repo memory banks.
get-shit-done
gsd-build/get-shit-done
A light-weight and powerful meta-prompting, context engineering and spec-driven development system for Claude Code by TÂCHES.
reddit-mcp-buddy
karanb192/reddit-mcp-buddy
Clean, LLM-optimized Reddit MCP server. Browse posts, search content, analyze users. No fluff, just Reddit data.