dsh-ponytail
Port of ponytail: an always-on lazy-senior-developer coding persona with intensity levels, persistence of the default level, and review, audit, debt, gain and help skills for DeepSeek Harness.
ponytail 移植：常驻的「懒惰资深开发者」编码人设，带强度档位与 review、audit、debt、gain、help 技能，适用于 DeepSeek Harness。
How to install
dsh plugin add @mengyuly/dsh-ponytail About
dsh-ponytail 把 DietrichGebert/ponytail（「懒惰资深开发者」最少代码心智）移植成 DeepSeek Harness 原生插件。功能与效率与上游一致：7 级阶梯规则集每轮注入、强度切换、/ponytail-* 斜杠命令。 安装 装进某个 profile（web 可换成 tui/自定义名）： bash 方式一：本地链接（当前 dsh 核 ≥ 0.1.x） dsh plugin --profile web add link:$(pwd) 方式二：GitHub 直接装 dsh plugin --profile web add github:MengYuil/dsh-ponytail 方式三：Release 打包件（先下 tgz） dsh plugin --profile web add file:./mengyuly-dsh-ponytail-0.1.3.tgz 方式四：npm dsh plugin --profile web add @mengyuly/dsh-ponytail 装完重启 profile 生效（dsh web / dsh tui）。装载完成后，会话技能目录里会出现 6 个 ponytail* 技能，发 /ponytail-help 立即验证。 lib/index.js 是自包含 bundle（已内联 dsh-llm / dsh-skil…
Recommendation signals
Meta
- License
- MIT
- Language
- JavaScript
- GitHub stars
- 1
- mo. downloads
- 63
- Last push
- 2026-08-24
- Created
- 2026-08-23
Basic safety check
- Findings
- None
- Sources
- curated:awesome-dsh-plugin.com, curated:awesome-dsh-plugin/awesome-dsh-plugin
- Topics
- ai-tools, deepseek, deepseek-harness, deepseek-harness-plugin, dsh, dsh-plugin, plugin, ponytail
Related plugins
dsh-awiki
AgentConnect/dsh-awiki
Provides DeepSeek Harness agents with native identities based on the open Agent Network Protocol (ANP), plus identity-based direct, group, mail, and Agent-to-Agent communication.
dsh-auth-everying
chenbin-dev/dsh-auth-everying
Import local Claude, Codex, Grok, Gemini, Copilot, OpenCode, and CC Switch config into DeepSeek Harness with OAuth login for supported providers.
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.
dsh-vision-router
ysr666/dsh-vision-router
Free vision for text-only agents: built-in keyless vision chain plus pixel tools (Q&A, grounding, crop, pixel diff, colors, OCR, SVG trace, cutout, screenshots); paste an image to use it.