Build a plugin once and use it across compatible agent clients.
Introducing Agent Plugins, an open standard developed with @awsdevelopers, @cursor_ai, @github, @code, and @vercel that packages Agent Skills and supports MCP server configurations in a shared format.
今天用 Kimi K3 给 CatReader 增加一广场的功能呢,其实改动量不少,包括前端后端和设计,差不多折腾了一天,上线了。Token 消耗比网传的少太多了。K3 的表现相当出色。
我觉得 Kimi 犯了一个宣传上的大问题,导致很多用户直接用 做 Coding 的事,但 Kimi App 做的是 Work 的事,用 Work 做 Coding,Token 消耗是急剧增加的。
我的 Vibe 群一堆用户用 Kimi App 而不是 Kimi Code cli 工具做编程,不烧 Token 才怪。
‼️FLOOD THE PHONES‼️
Tell the Senate to remain in DC until the SAVE America Act is passed.
Every Senate number begins w/ 202-224-XXXX.
“My name is ____. My zip code is ____. I’m asking Senator ____ to remain in DC until the SAVE America Act is passed. I vote in every election.”
Peacefully.
We tried using Meta's new Muse Code agent, but it has a bug that doesn't let it sign in from a docker container.
So we did a fun experiment: Meta claims Muse Spark 1.2 was co-trained with their Muse agent harness. So we extracted instructions from their system prompt and added them to the Cline harness.
TL;DR of this special prompting:
- Trust source code over the user prompt, so read every call site and existing tests before starting the task
- Weigh edge and error cases as heavily as the happy path
- Always reproduce the bug before fixing
- Don't trust the first passing test suite, and verify suspicious looking half-baked tests
- Never stop at just editing, keep working until the change is verified complete.
We then asked this modified harness to fix a real bug from our repo, and compared the results to the original Cline agent harness.
Results:
- Used 2.7x fewer tokens (19.7M → 7.2M)
- Finished 2x faster (49min → 24min)
- Cost 2.4x less ($7.69 → $3.25)
Same Muse Spark 1.2 model, same task, only the prompting changed. Incredible how much of a performance gain Meta was able to achieve training it on these special instructions!