akqwpeter-prog/skill-bartender0

skill-bartender

具有延迟阶梯和安全安装仓库的“任务-技能”配对工具。绝不超量注入,绝不提供未测试的技能。支持 DSH、Claude Code、Codex

AI 分析

核心用途是优化 Agent 的技能加载策略,防止其加载过多、错误或遗漏工作流技能。适合需要精准、安全地进行任务与技能配对的 DSH、Claude Code 或 Codex 用户。

包名
skill-bartender
版本
0.1.1
许可证
MIT
最近更新
2026年8月17日

安装

$npx -p @deepseek-ai/dsh dsh plugin --profile web add github:akqwpeter-prog/skill-bartender

skill-bartender 🍸

License SkillSpector CI dsh-plugin

Task-to-skill pairing with a laziness ladder and a safe install cellar.

Your agent already sees a catalog of skill names and descriptions — but it over-pours: loads too many skills, loads the wrong ones, or misses the one workflow skill that composes the task. skill-bartender is the meta-skill that fixes the pour:

  • 🪜 Laziness ladder — zero skills when plain tools suffice; one skill when one matches; workflow over hand-composed atomics; unsure → don't load. (Spirit: ponytail — the best load is the load never made.)
  • 🍷 Routing table — a user-editable task→skill map (references/policy.md).
  • 🔐 Safe cellar — when a needed skill is missing: quarantine → SkillSpector scan → explicit human approval → copy into the skills dir. Never auto-installs.
  • 🧠 Learn — loaded-but-unused skills get logged and skipped next time.

Install (one file, three platforms)

# DeepSeek Harness
mkdir -p ~/.dsh/skills/skill-bartender
cp skills/skill-bartender/SKILL.md ~/.dsh/skills/skill-bartender/
cp -r skills/skill-bartender/references ~/.dsh/skills/skill-bartender/

# Claude Code
mkdir -p ~/.claude/skills/skill-bartender
cp skills/skill-bartender/SKILL.md ~/.claude/skills/skill-bartender/

# Codex
mkdir -p ~/.codex/skills/skill-bartender
cp skills/skill-bartender/SKILL.md ~/.codex/skills/skill-bartender/

Then say "skill-bartender" once, or paste the routing table into your AGENTS.md for always-on routing.

Or install as a DeepSeek Harness bundle:

dsh plugin --profile web add github:akqwpeter-prog/skill-bartender

Self-scan

This skill scans itself clean: SkillSpector 0 findings (score 0 / SAFE). Report: docs/skillspector-report.json.

Security model (read this)

  • Skills are instructions, and instructions can be adversarial (prompt injection). SkillSpector is a filter, not a guarantee.
  • scripts/ in any skill is code — never executed without human review.
  • Human approval is mandatory for every install. No silent installs, ever.

Not reinventing the wheel

Use skill-bartender alongside the ecosystem, not instead of it: DshMarket (GUI marketplace), dsh-find-plugin (conversational search), dsh-plugin-autoevo (plugin auto-install with approval). skill-bartender adds the routing policy and the quarantine-then-approve discipline for skills.

Eval

The routing policy ships with a gold-task suite: docs/eval.md.

License

MIT. Ponytail (MIT) is referenced, not bundled — tribute in the SKILL.md.