展示HN：Skrun – 将任何代理技能部署为API

展示HN：Skrun – 将任何代理技能部署为API
Show HN: Skrun – Deploy any agent skill as an API

原始链接: https://github.com/skrun-dev/skrun

## Skrun：将代理技能转化为API Skrun是一个开源工具，用于部署和调用AI代理技能作为API。它允许你轻松地将指令（定义在`SKILL.md`中）和配置（`agent.yaml`）转化为可通过`POST /run`端点调用的服务。 **主要特性：** * **简单部署：** 使用`skrun init`创建新的代理或导入现有技能，然后使用`skrun deploy`使其可通过API访问。 * **多模型支持：** 支持Anthropic、OpenAI、Google、Mistral和Groq，并具有自动回退功能。 * **有状态代理：** 代理可以使用键值状态在多次运行中记住信息。 * **工具调用：** 支持本地CLI工具和MCP服务器，以扩展功能。 * **类型化I/O：** 定义结构化输入和输出，以实现可靠的API交互。 Skrun提供了一个本地开发服务器（`skrun dev`）用于测试，以及一个CLI用于管理代理、构建和部署。示例展示了各种应用，例如代码审查、PDF处理和数据分析。云部署正在计划中，基于现成的架构。

原文

Turn any Agent Skill (SKILL.md) into a callable API via POST /run. Multi-model, stateful, open source.

npm install -g @skrun-dev/cli

# Import an existing skill → deploy → call
skrun init --from-skill ./my-skill
skrun deploy

curl -X POST localhost:4000/api/agents/dev/my-skill/run \
  -H "Authorization: Bearer <token>" \
  -H "Content-Type: application/json" \
  -d '{"input": {"query": "analyze this"}}'

skrun init my-agent
cd my-agent
# Creates SKILL.md (instructions) + agent.yaml (config)

skrun init --from-skill ./path-to-skill
# Reads SKILL.md, asks 2-3 questions, generates agent.yaml

skrun dev
# ✓ Server running at http://localhost:3000
# POST /run ready — watching for changes...

skrun test
# ✓ basic-test (output.score >= 0)
# 1 passed, 0 failed

skrun deploy
# ✓ Validated → Built → Pushed
# 🚀 POST http://localhost:4000/api/agents/you/my-agent/run

v0.1 ships with a local runtime. Cloud deploy is on the roadmap — the architecture is ready (RuntimeAdapter interface).

Agent Skills — SKILL.md standard, compatible with Claude Code, Copilot, Codex
agent.yaml — Runtime config: model, inputs/outputs, permissions, state, tests
POST /run — Every agent is an API. Typed inputs, structured outputs.
Multi-model — Anthropic, OpenAI, Google, Mistral, Groq with automatic fallback
Stateful — Agents remember across runs via key-value state
Tool calling — Two approaches: CLI tools (scripts/ — write your own, bundled with the agent) and MCP servers (npx — standard ecosystem, same as Claude Desktop)

All examples use Google Gemini Flash by default. Change the model section in agent.yaml to use any supported provider.

Agent	What it shows
code-review	Import a skill, get a code quality API
pdf-processing	Tool calling with local scripts
seo-audit	Stateful — run twice, it remembers and compares
data-analyst	Typed I/O — CSV in, structured insights out
email-drafter	Business use case — non-dev API consumer
web-scraper	MCP server — headless browser via @playwright/mcp

# 1. Start the registry
cp .env.example .env          # add your GOOGLE_API_KEY
pnpm dev:registry              # keep this terminal open

# 2. In another terminal
skrun login --token dev-token
cd examples/code-review
skrun build && skrun push

# 3. Call the agent
curl -X POST http://localhost:4000/api/agents/dev/code-review/run \
  -H "Authorization: Bearer dev-token" \
  -H "Content-Type: application/json" \
  -d '{"input": {"code": "function add(a,b) { return a + b; }"}}'

Windows (PowerShell): use curl.exe instead of curl, and use @input.json for the body.

Command	Description
`skrun init [dir]`	Create a new agent
`skrun init --from-skill <path>`	Import existing skill
`skrun dev`	Local server with POST /run
`skrun test`	Run agent tests
`skrun build`	Package .agent bundle
`skrun deploy`	Build + push + live URL
`skrun push` / `pull`	Registry upload/download
`skrun login` / `logout`	Authentication
`skrun logs <agent>`	Execution logs

Full CLI reference →

git clone https://github.com/skrun-dev/skrun.git
cd skrun
pnpm install && pnpm build && pnpm test

See CONTRIBUTING.md for conventions and setup.

MIT

展示HN：Skrun – 将任何代理技能部署为API Show HN: Skrun – Deploy any agent skill as an API

展示HN：Skrun – 将任何代理技能部署为API
Show HN: Skrun – Deploy any agent skill as an API