shopkeeper-agent
📊 电商数仓智能问数 AI Agent,最适合用于系统学习 LangGraph 的实战项目:基于 LangGraph、FastAPI、Qdrant、Elasticsearch、MySQL 与 React,完整实现元数据知识库、混合检索、自然语言生成 NL2SQL 生成校验、SQL 执行与流式查询展示。前后端完整代码全栈可跑,Docker 环境一键部署,配套 ai-agents-from-zero 免费教程与章节代码分支。适合系统学习大模型应用、数据分析 Agent 和企业级 AI 工程落地。
Runs with mockssource: GitHubhomepagePythonMITcommit 8045fa4d6160
Python, MIT licensed. The project labels itself: agent workflow, ai agent, elasticsearch, fastapi, langchain, langgraph, llm and nl2sql.
shopkeeper-agent runs, with stand-ins for the services it depends on. An Argusic agent installed it in 28 minutes and hit 4 errors and fixed every one without help on a clean machine with no GPU, and the whole session was recorded.
With schema-correct mocks for the mandatory external services, the uvicorn-served app answers POST /api/query with a complete SSE pipeline and returns SQL aggregate result sales=10088 for '统计华北地区销售额'; uv sync installs all dependencies and the frontend pnpm build completes.
What the agent ended up with on a clean machine, in its own words. How this is measured
Measured by Argusic on a fresh machine every time. Every number links to its evidence. Argusic Score 92 of 100 (the mean of this project's run scores).
At a glance
Subject data from GitHub, linked at the top of this page, refreshed . Test data by Argusic (CC BY 4.0); every number links to a run page with the full log, the recording, and their sha256 hashes.
What was measured
From the run that decided this verdict. Every line is a recorded value, not an estimate.

What went wrong, in the agent's own words
The agent's own words, worst first, never rewritten into advice.
- App lifespan requires external MySQL/Qdrant/ES/TEI/LLM not present in container20 minutes
- pip install -e . fails: setuptools flat-layout (app, conf, docker, prompts, frontend)3 minutes
- Manual pip venv (langchain-huggingface 1.2.3) rejects embedding model=URL; uv pin 1.2.1 accepts it3 minutes
- Config requires LLM_API_KEY; omegaconf.InterpolationResolutionError without it2 minutes
- Install time
- 28 minutes
- Cold machine to finish
- 31 minutes
- Errors hit and fixed
- 4 hit, 4 fixed with no human help
- How the result was proved
- uv run pytest tests/test_integration.py -v exited 0 with 1 passed (asserts /api/query 200 and SSE pipeline ends with result data [{'sales':10088}]); also curl -X POST http://127.0.0.1:8000/api/query returned HTTP 200 with full SSE stream ending in [{'sales':10088}]
- Model tokens used
- 497,338
- Exact commit tested
- 8045fa4d6160
- Container image
- sha256:cdd920bce783
What was not measured
The other half, so you can see exactly how far this evidence goes. A gap is printed as a gap and never filled in.
- Peak GPU memory
- this ran on a CPU machine, so there was no GPU to sample
- Behaviour over a long run
- each attempt uses the smallest documented settings and a fixed time budget, so this measures whether it runs, not how it holds up under load
- Behaviour against real services
- this run was verified against stand-in services only
The end of the recording
The last lines of the terminal session that decided this verdict, exactly as they were recorded.
- App imports under `uv run` (`import main` and `uv run fastapi dev main.py` path).
**Errors encountered and fixed**
1. Config fails without the LLM key (`omegaconf.InterpolationResolutionError: 'LLM_API_KEY' not fou…
2. `pip install -e .` (non-documented path) fails with setuptools flat-layout error (multiple top-l…
3. App lifespan requires external MySQL, Qdrant, Elasticsearch, TEI embedding, and an LLM — none pr…
4. A dependency-version nuance: in a pip venv with `langchain-huggingface 1.2.3`, the embedding `mo…
**Verification**
- `uv run pytest tests/test_integration.py -v` → **1 passed**.
- The test launches the app lifespan and POSTs to `/api/query` with `{"query":"统计华北地区销售额"}`; the SS…
- I also launched the real `uvicorn` (with mocks) and `curl -X POST http://127.0.0.1:8000/api/query…
**Final state**: dependencies install via `uv sync`; the app builds and runs; `/api/query` executes…
Notes on remaining work: the repo has no unit tests originally; I added `tests/test_integration.py`…Replay the whole session, every command from a clean machine to this point.
How it was tested
One agent, one container, no human help. It installs the project the way its own documentation says to, runs it, and fixes what breaks. Everything below is recorded as it happened: the terminal session, the log and the exact commit. The full method.
Strengths and limits
Measured facts, not opinions. How this is written.
What went well
- Installed in 28 minutes.
- Recovered from all 4 errors without a human stepping in, which says the failures are documented well enough to solve.
- Licensed MIT, as reported by its host.
What did not, or is not known
- Only reached a running state against stand-in services, so nothing here proves it works against the real ones.
- Measured once so far. A single attempt shows it can run, not that it runs reliably.
Also tested, in the same area
Every one of these was installed and run by Argusic on a clean machine. Nothing appears here that was not tested.
Run history
| tested by | attempt | outcome | score | cost USD | when |
|---|---|---|---|---|---|
| Argusic Runner | 1/3 | Runs with mocks | 92.00 | 0.59 |
Topics (from GitHub)
agent-workflowai-agentelasticsearchfastapilangchainlanggraphllmnl2sqlqdrantragvector-database
Embed the badge
Markdown for the project README. It links back here; terms on the terms page.
[](https://argusic.com/subject/shopkeeper-agent)Questions
- Does shopkeeper-agent run?
- shopkeeper-agent runs, with mocks standing in for external services. Argusic installed and launched it on a clean machine in 28 minutes, hitting 4 errors on the way, and recorded the session.
- How did Argusic test shopkeeper-agent?
- On a fresh, disposable machine, with every command recorded and the repository pinned at commit 8045fa4d6160. One attempt is recorded, and the full method is on the methodology page.
- What did it cost to test shopkeeper-agent?
- The run that produced this verdict cost $0.59: real compute and model cost, not a list price per million tokens. It is recorded on that run's page.
- How long does shopkeeper-agent take to install?
- 28 minutes in the run that decided this verdict, from a clean machine to a running state. That is one measured attempt, not an average.
- Does shopkeeper-agent need a GPU?
- Not to start. Argusic reached a running state on a machine with no GPU.
- What goes wrong when installing shopkeeper-agent?
- 4 things broke in the recorded run, and 4 were fixed without human help. Each one, and the time it cost, is listed on this page.
- Where is the evidence for shopkeeper-agent?
- The recorded run is on this page, each linking to its full log and terminal recording, stored with a sha256 fingerprint so it cannot be quietly altered.