围绕「Show HN: SwarmWright, structured multi-agent AI defined in markdowns」：从 agent/MCP/harness/成本与工程化视角抽取可迁移结论，并落到可执行动作。

为什么重要

收藏=处理集；无 ka 即断档。Mode B 1:1 补回 star→文章链路。

正文

围绕「Show HN: SwarmWright, structured multi-agent AI defined in markdowns」：从 agent/MCP/harness/成本与工程化视角抽取可迁移结论，并落到可执行动作。

Claim

Why it matters

收藏=处理集；无 ka 即断档。Mode B 1:1 补回 star→文章链路。

Summary

Mode B 回补：FR favorite + Vault 无 ka-v0。标题：Show HN: SwarmWright, structured multi-agent AI defined in markdowns

Actions

(none)

Evidence

(none)

Caveats

(none)

Research queries

(none)

Body

正文

背景

FR 收藏条目《Show HN: SwarmWright, structured multi-agent AI defined in markdowns》曾在批量清未读时只 mark_read、未走沉淀；Mode B 从 favorite 无 ka 队列回补。

机制

原文：https://www.swarmwright.com/
收藏 id：1778879738992755
从标题/摘要与领域关键词提炼 claim；evidence 至少含原文 + 公开相关页。

取舍

优先可迁移工程动作，避免八卦/额度刷屏；弱信号走 shallow。

动作

打开原文核对；把 claim 并入 harness/MCP/成本基线；需要时再 Exa 加厚杂志长文。

结合的源文章

主源

Show HN: SwarmWright, structured multi-agent AI defined in markdowns

打开原文 ↗

原文快照

展开 / 收起快照

I had a bunch of custom AI pipelines and a growing folder of markdown files and Python scripts holding it together. Built this to give that chaos some structure.

Agents are markdown files, topology is a JSON file the runtime enforces hard. The agents are still fully autonomous: they make their own decisions, but the graph they operate in isn't. You declare who can call whom upfront and the runtime holds that line.

No auth yet, fine if you don't expose the port, i guess. Two Docker commands to run it.

Comments URL: https://news.ycombinator.com/item?id=48153623

Points: 3

# Comments: 0