通过 Bohrium 开放平台代理进行网络搜索(由 searchapi.io 支持)。适用场景:用户需在开放网络上搜索研究论文、文档、教程等。
Proxy to searchapi.io via open.bohrium.com's /v1/search/web endpoint. Issues keyword queries against the open internet and returns a list of hits with title, URL, and snippet.
Use when: locating a software's homepage, a blog post, a quick fact check, a news article.
Don't use for:
bohrium-paper-search or bohrium-scholar-searchbohrium-knowledge-baseNo CLI support — HTTP API only. The bohr CLI ships a built-in BohrWebSearch tool that calls this endpoint automatically.
ACCESS_KEY comes from the OpenClaw config file ~/.openclaw/openclaw.json:
json
"bohrium-web-search": {
"enabled": true,
"apiKey": "YOUR_ACCESS_KEY",
"env": {
"ACCESS_KEY": "YOUR_ACCESS_KEY"
}
}
OpenClaw injects env.ACCESS_KEY into the runtime.
text
GET https://open.bohrium.com/openapi/v1/search/web?q=QUERY&num=N
Header: accessKey: $ACCESS_KEY
ParamTypeDefaultDescriptionqstringrequiredSearch querynumint3Number of results, range 1-10
python
import os, requests
AK = os.environ["ACCESS_KEY"]
BASE = "https://open.bohrium.com/openapi/v1/search/web"
r = requests.get(BASE,
headers={"accessKey": AK},
params={"q": "graphene synthesis CVD", "num": 5})
data = r.json()
for i, hit in enumerate(data.get("organic_results", []), 1):
print(f"[{i}] {hit['title']}")
print(f" {hit['link']}")
print(f" {hit.get('snippet', '')[:200]}")
print()
Response fields:
FieldDescriptionorganic_resultsArray of main resultsorganic_results[].titlePage titleorganic_results[].linkPage URLorganic_results[].snippetExcerptorganic_results[].positionResult rank
bash
AK="YOUR_ACCESS_KEY"
curl -s "https://open.bohrium.com/openapi/v1/search/web?q=deepmd-kit&num=5" \
-H "accessKey: $AK" | jq '.organic_results[] | {title, link, snippet}'
SymptomCauseFixNo organic_resultsNo matches for the queryRephrase; English queries generally return more hits than Chinese401Bad ACCESS_KEYCheck accessKey header case; don't use Authorization: Bearer``num ignoredOut of rangeMust be 1-10; values outside the range may be clamped or ignored