Skip to content

drag the cards

Shipping agents into real products

field notes, written in Chinese and English

Liu ZhuoQi · AI Application Engineer · Agent Systems

Pencil sketch portrait of Liu ZhuoQi — me, most days ~

Outcomes is only as good as the rubric you write.

— from Agents That Check Their Own Work and Tidy Their Own Memory

zhuoqi@dev — zsh

$ whoami
Liu ZhuoQi · AI application engineer (agents)

$ ls posts/ | wc -l
18  # every post also in Chinese

$ cat focus.md
agent memory · grading & review · inference
Temporal workflows · shipping to prod

$ open latest.md

Latest

Check the Goods Before You Buy a Relay API

Cheap Claude, GPT and Seedance relays keep multiplying, and their groups are labelled 'official', 'official key' or 'native', but a group name is not evidence. What actually ties what a relay sells to the official API is the response id prefix, input tokens that match the official counting endpoint one for one, the error you get when you send a deliberately invalid parameter, the C2PA signature and signing certificate inside a video file, and the charges in the relay's own logs. This post walks through six questions: how to check each one, whether it costs money, what counts as a problem, and what to do without an official key. It ends with an open-source skill that hands the whole procedure to an Agent and writes the comparison report for you.

2026-10-24Read →

series/

Agent Architecture Deep Dives 4 posts

series/

OpenClaw Production Notes 3 posts

Engineering · 05-16

Why We Moved from Celery to Temporal

Agent pipelines have state, get stuck and need replay; one failure must not sink the batch.

Field guide · 07-19

How to Choose an LLM Inference Engine

From a single local GPU to PD disaggregation: 8 engines in 3 tiers, with a decision tree.

Deep dive · 07-30

How Agents Remember You

From Ebbinghaus to engrams, then a code audit of Mem0, Letta, Graphiti and three more.

Projects

AI vendor statusWhat the official status pages of OpenAI, Claude, Gemini, xAI, Mistral, Groq, Amazon Bedrock, DeepSeek, Kimi, MiniMax and others say right now, set as a newspaper front page — the worst one makes the headline, every vendor gets a weather, open one for components and incidents in progress; vendors without a status page are listed too, with the reason. A Cloudflare Pages Function reads each vendor at most once a minute and never calls a model.llm-relay-auditAn open-source Agent skill that compares an AI API relay (new-api / one-api / sub2api) with the official API item by item: real model or swapped, injected prompts, official original video, swallowed parameters, real charges. Every raw request is logged, and it ends with an external evidence PDF and an internal decision card. Python + the Agent Skills spec, one-line npx install.ArgusAn open-source, frontend-only agent harness for long-video understanding: the video file never leaves the browser, multi-provider LLMs via Vercel AI SDK, with frame-extraction / memory / sub-agent tools and dual-CDN publishing. Vite + React + TypeScript.VaneA self-built AI Agent system that ran in production from 2026-07 to 09 and is now paused: multi-source ingestion → LLM personalized scoring → Feishu push, with a user-profile feedback loop. Go + Temporal + A2A protocol + React, full-stack.

CKACKSRHCE