SKILL: Bohrium Scholar Search & Profile#

Overview#

Query scholar information via the Bohrium OpenAPI gateway ( open.bohrium.com). Provides two core endpoints:

EndpointMethodPathPurposeScholar SearchPOST/openapi/v1/paper-server/scholar/searchSearch scholars by name / affiliation / research tagsScholar InfoGET/openapi/v1/paper-server/scholar/info?scholarId=xxxFetch the full profile by scholarId

Typical workflow: Given a scholar name → search candidates → pick the target scholarId → fetch the full profile (papers, citations, h-index, research directions, education/work history).

No CLI support — all operations use the HTTP API.

Authentication#

ACCESS_KEY is read from the OpenClaw config ~/.openclaw/openclaw.json:

json

"bohrium-scholar-search": {
  "enabled": true,
  "apiKey": "YOUR_ACCESS_KEY",
  "env": {
    "ACCESS_KEY": "YOUR_ACCESS_KEY"
  }
}

OpenClaw automatically injects env.ACCESS_KEY into the runtime environment.

Runtime Resolution#

text

Read os.environ["ACCESS_KEY"]
  ├─ Non-empty → use directly
  └─ Empty     → prompt the user:
                 "ACCESS_KEY is not configured in OpenClaw. Please set
                  bohrium-scholar-search.env.ACCESS_KEY in ~/.openclaw/openclaw.json
                  with the AccessKey obtained from the user settings page at
                  https://bohrium.dp.tech, then restart the OpenClaw session."

Important: Do not persist the AccessKey in any other file or hardcode it into source; always inject it through the OpenClaw environment.

Error Handling#

If the API returns Invalid AccessKey (code 2000) or HTTP 401:

1. The key configured in OpenClaw is invalid or expired.
2. Prompt the user: "Your AccessKey is invalid. Please update bohrium-scholar-search.env.ACCESS_KEY in ~/.openclaw/openclaw.json and restart the OpenClaw session."

Common Code Template#

python

import os, requests

AK = os.environ.get("ACCESS_KEY", "")
if not AK:
    raise RuntimeError(
        "ACCESS_KEY not found. Please configure it in ~/.openclaw/openclaw.json "
        "under bohrium-scholar-search.env.ACCESS_KEY."
    )

BASE = "https://open.bohrium.com/openapi/v1/paper-server"
HEADERS_JSON = {"accessKey": AK, "Content-Type": "application/json"}
HEADERS = {"accessKey": AK}

---

Standard Workflow#

text

User asks about a scholar
  │
  ├─ Known scholarId → call Scholar Info directly
  └─ Unknown scholarId → call Scholar Search first
       └─ Pick items[].scholarId from the response → call Scholar Info

---

Scholar Search#

Basic Search#

python

r = requests.post(f"{BASE}/scholar/search", headers=HEADERS_JSON, json={
    "name": "Yann LeCun",
    "page": 1,
    "pageSize": 5
})
data = r.json()
for item in data["data"]["items"]:
    print(f"[{item['scholarId']}] {item.get('nameEn','')} / {item.get('nameZh','')}")
    print(f"  Org: {item.get('scholarOrgNameEn','')}")
    print(f"  Papers: {item.get('paperNums',0)}, Citations: {item.get('citationNums',0)}, h-index: {item.get('hIndex',0)}")

Search with Filters#

python

r = requests.post(f"{BASE}/scholar/search", headers=HEADERS_JSON, json={
    "name": "Zhang San",
    "school": "Tsinghua University",
    "affiliation": "Tsinghua University",
    "tags": "machine learning",
    "page": 1,
    "pageSize": 10
})

Request Parameters#

ParameterTypeRequiredDescriptionnamestringYesScholar name keyword (1–99 chars)schoolstringNoSchool / institutiontagsstringNoResearch interest tagaffiliationstringNoAffiliation (English)affiliationZhstringNoAffiliation (Chinese)pageintNoPage number, default 1pageSizeintNoPage size, default 10

Key Response Fields#

data.items[] array, each item contains:

FieldDescriptionscholarIdUnique scholar ID, required by the info endpointnameEn / nameZhEnglish / Chinese namepaperNumsPublication countcitationNumsCitation counthIndexh-indexscholarOrgNameEn / scholarOrgNameZhAffiliation name (EN/ZH)isHighCitedHighly-cited scholar flag

---

Scholar Info#

Fetch the full profile using the scholarId returned by the search endpoint:

python

r = requests.get(
    f"{BASE}/scholar/info",
    headers=HEADERS,
    params={"scholarId": scholar_id}
)
info = r.json()["data"]
print(info.get("nameEn"), "|", info.get("nameZh"))
print("Research:", info.get("researchDirection"))
print("Education:", info.get("educationBackgroundZh") or info.get("educationBackground"))
print("Work:", info.get("workExperienceZh") or info.get("workExperience"))

Additional Response Fields#

In addition to the fields returned by search:

FieldDescriptionresearchDirectionArray of research directionseducationBackground / educationBackgroundZhEducation history (EN/ZH)workExperience / workExperienceZhWork experience (EN/ZH)

---

Presentation Guidelines#

Format the API response into a user-friendly summary that highlights:

• Name: nameEn / nameZh
• Affiliation: scholarOrgNameEn / scholarOrgNameZh
• Metrics: paperNums / citationNums / hIndex
• Highly cited: isHighCited
• Research directions: researchDirection
• Education: prefer educationBackgroundZh, fall back to educationBackground
• Work experience: prefer workExperienceZh, fall back to workExperience

If multiple candidates are returned, first present a summary table for the user to choose, then fetch the full profile.

---

curl Examples#

bash

AK="$ACCESS_KEY"
BASE="https://open.bohrium.com/openapi/v1/paper-server"

# Step 1: Scholar search
curl -s -X POST "$BASE/scholar/search" \
  -H "accessKey: $AK" \
  -H "Content-Type: application/json" \
  -d '{"name":"Yann LeCun","page":1,"pageSize":3}'

# Step 2: Fetch profile
curl -s "$BASE/scholar/info?scholarId=RETURNED_ID" \
  -H "accessKey: $AK"

---

Troubleshooting#

ProblemCauseSolutionACCESS_KEY is emptyOpenClaw did not inject the env varVerify bohrium-scholar-search.env.ACCESS_KEY in ~/.openclaw/openclaw.json``Invalid AccessKey / 401Key expired or incorrectUpdate the AccessKey in ~/.openclaw/openclaw.json and restart the sessionEmpty search result1) A 24-char no-space string is treated as an internal ID; 2) upstream per-user rate limitingUse a natural name rather than an ID; retry laterInfo endpoint parameter errorMissing scholarIdCall the search endpoint first and take data.items[].scholarId``name length errorOut of 1–99 charsShorten the name keyword