argo
专门为 agent 打造的开源 agent 搜索工具，具备 200+的搜索源及多语言搜索能力，覆盖中文/英文/学术/代码/购物/金融/新闻/百科。
Runssource: GitHubPythonApache-2.0commit 32428bc90aae
Python, Apache-2.0 licensed. The project labels itself: academic search, agent tools, ai agent, deep research, dsh ecosystem, dsh plugin, evidence verification and mcp server.
argo runs. An Argusic agent installed it in 65 minutes and hit 11 errors on a clean machine with no GPU, and the whole session was recorded.
With a local venv and deps installed, the argo CLI (v2.9.7) runs search/fetch and returns real results, the MCP stdio server answers argo_search call, headless Chrome renders pages, and the full test suite passes 4258 tests with 36 platform-gated skips and 0 failures.
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 98.2 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.
- test_fetch_v3_identity_branch/test_tinyfish_fallback/test_evidence_loop failed: 'blocked' due to SSRF guard on unresolvable fixture domains (offline DNS can't resolve v.com/docs.example.com)15 minutes
- test_funnel/test_race_fast_budget/test_lowlevel_smoke failed after installing rapidfuzz: fuzzy near-dup layer auto-enabled and collapsed template rows differing only by index12 minutes
- test_fallback_extract/test_local_360_jisilu/test_local_baidu_pipeline/test_local_search_redirect_resolve/test_serp_guard failed: no BeautifulSoup (bs4)10 minutes
- test_cdp_network_idle/test_chrome_leak_fix failed: FileNotFoundError Chrome not found8 minutes
- test_platform_paths_conventions env-cache test failed: overlayfs keeps mtime_ns unchanged on equal-size back-to-back rewrite, cache returned stale value6 minutes
- test_local_image 5 tests failed: FileNotFoundError 'swiftc' (Vision helper needs macOS)5 minutes
- Install time
- 65 minutes
- Cold machine to finish
- 60 minutes
- Errors hit and fixed
- 11 hit, 10 fixed with no human help
- How the result was proved
- PATH=$HOME/.local/bin:$PATH CHROME_PATH=$HOME/.cache/ms-playwright/chromium-1248/chrome-linux64/chrome python3 -m pytest tests/ -q -> 4258 passed, 36 skipped, 57 subtests passed, 0 failed in 374.41s; `argo --version` -> 'argo 2.9.7'; `argo search 'python asyncio tutorial' --json -n 2 --no-cache` -> status=completed count=2 errors=[] first_url=https://itsourcecode.com/blogs/python-async-await-tutorial-2026; `argo fetch https://example.com --json` -> success:true fetch_method:http length:156; `python3 scripts/mcp_server.py --call argo_search '{"query":"python asyncio","max_results":2}'` -> JSON results rc=0; headless ChromeCDP navigate example.com -> title 'Example Domain' on DISPLAY=:99
- Model tokens used
- 1,207,119
- Exact commit tested
- 32428bc90aae
- 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
The end of the recording
The last lines of the terminal session that decided this verdict, exactly as they were recorded.
**Verified**
- `argo --version` printed `argo 2.9.7`.
- `argo search "python asyncio tutorial" --json -n 2 --no-cache` returned `status=completed count=2…
- `argo fetch https://example.com --json` returned `success:true fetch_method:http length:156`.
- MCP `python3 scripts/mcp_server.py --call argo_search '{"query":"python asyncio","max_results":2}…
- Headless Chrome (via `ChromeCDP`) navigated `https://example.com` and read `document.title == "Ex…
- `python3 -m pytest tests/ -q` (with `PATH` including `~/.local/bin` and `CHROME_PATH` set): **425…
- `python3 scripts/health_check.py --all` returned `0/18 可用` — local_* engines are blocked by anti-…
**Unfixed gaps**
- `scripts/health_check.py` local-engine probes (local_baidu/sogou/360/bing_news/wikipedia) all fai…
- Local image Vision (Swift) needs macOS; skipped on Linux by design.
- Optional third-party keys (OCTEN_API_KEY etc.) are absent; engine auto-skips them and no mock was…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
- Reached a running state on a clean machine, with the session recorded.
- Installed in 65 minutes.
- Ran without a GPU, so it does not need one to start.
- Licensed Apache-2.0, as reported by its host.
What did not, or is not known
- Hit 11 errors during setup, 1 of which were not solved.
- Measured once so far. A single attempt shows it can run, not that it runs reliably.
What it is a reasonable choice for
- Trying it on a laptop or a small server: it reached a running state without a GPU.
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
Topics (from GitHub)
academic-searchagent-toolsai-agentdeep-researchdsh-ecosystemdsh-pluginevidence-verificationmcp-servermetasearchpythonsearch-engineweb-search
Embed the badge
Markdown for the project README. It links back here; terms on the terms page.
[](https://argusic.com/subject/argo)Questions
- Does argo run?
- Yes. argo runs. Argusic installed and launched it on a clean machine in 65 minutes, hitting 11 errors on the way, and recorded the session.
- How did Argusic test argo?
- On a fresh, disposable machine, with every command recorded and the repository pinned at commit 32428bc90aae. One attempt is recorded, and the full method is on the methodology page.
- What did it cost to test argo?
- The run that produced this verdict cost $0.77: real compute and model cost, not a list price per million tokens. It is recorded on that run's page.
- How long does argo take to install?
- 65 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 argo need a GPU?
- Not to start. Argusic reached a running state on a machine with no GPU.
- What goes wrong when installing argo?
- 11 things broke in the recorded run, and 10 were fixed without human help. Each one, and the time it cost, is listed on this page.
- Where is the evidence for argo?
- 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.