Agent skill · Audited from source
Academic Literature Search
skill · InternLM/WildClawBenchMIT licenceAudited 15 Oct 2026 · source a79d9cd3663a
Is it safe? Yes — SAFE, grade B (89/100). We read this skill's source at commit a79d9cd3663a and observed network access and no hard-coded secrets. What it does: An in-the-wild benchmark for AI agents in the production harness.
Cite
Cite this page
ModelsAtlas. “Academic Literature Search (InternLM/WildClawBench)”. Retrieved 15 Oct 2026. https://modelsatlas.com/skills/internlm/academic-literature-search
Trust audit
Audit 0.4.1 · source read 15 Oct 2026
SAFE means nothing in the source contradicts what the skill says it does. Grade B is shared by 16,667 of 17,602 audited skills.
What the source does
- FilesystemNone observedNo reads or writes in the source
- NetworkDeclared1 observation
- ShellNone observedNo commands run in the source
- DependenciesNot pinnedSome versions float
- SecretsNone foundNo keys in source
Five audit layers
| Layer | What we check | Result | Findings |
|---|---|---|---|
L0 Provenance | Repo owner, licence, archive status, activity | PASS | 0 |
L1 Static source | Secrets, path traversal, supply chain, hidden code | PASS | 1 |
L2 Instructions | Prompt injection, authority framing, hidden directives | PASS | 0 |
L3 Inventory | The files and declarations match the source | PASS | 1 |
L4 Sandbox | Run in isolation and watch behaviour | SKIPPED why | 0 |
Gates applied: no behavioural pass.
Findings (2)
LOWSupply chainsupply.unpinned · requirements.txt
requests, aiohttp, pandas, pyyaml, aiofiles
- Why
- 5 requirement(s) not pinned with ==
- Fix
- pin exact versions
LOWInventory / provenanceskill.no_frontmatter · SKILL.md:1
- Why
- SKILL.md lacks name/description frontmatter
- Fix
- —
Audit 0.4.1 · source sha a79d9cd3663a · 15 Oct 2026Report an issue or request a re-scan
Provenance
Who published it, and what the verdict is pinned to
- PublisherInternLM GitHub organizationOwner of the repository this skill is read from
- Source folderInternLM/WildClawBench/skills/4/academic-literature-searchThe exact folder we audited
- LicenceMITAs the repository states it
- Commit pinned for this verdicta79d9cd3663aA new audit runs when the source changes
- Repository statusNot archivedLast push 27 days ago
- Sandboxed runNot runSkills are read, not executed; layer L4 is skipped
What it tells the agent
Full source on GitHub ↗Its description: An in-the-wild benchmark for AI agents in the production harness.
# 学术文献检索技能
## 概述
这是一个专注于学术文献检索的专业工具,集成了多个权威学术数据库,提供全面、快速、准确的文献检索服务。支持多数据库并发检索、高级过滤、智能排序和多种输出格式。
## 核心功能
### 🔍 强大的检索能力
- **多数据库集成**:Semantic Scholar、Crossref、arXiv、PubMed
- **智能查询解析**:自然语言、布尔运算、字段限定、短语搜索
- **并发检索**:同时查询多个数据库,毫秒级响应
- **高级过滤**:年份、引用数、期刊类型、开放获取、语言等
### 📖 高级检索特性
- **自然语言查询**
- **布尔运算符** (AND, OR, NOT)
- **字段限定搜索** (title:, author:, year:)
- **范围搜索** (year:2020-2024, citations:>100)
- **通配符搜索**
### 🎯 精准的结果处理
- **智能去重**:基于DOI、标题、作者等多维度去重
- **多维度排序**:引用数、年份、相关性、影响力、趋势
- **高级过滤**:(期刊、开放获取、文献类型)
- **结果丰富**:自动补充元数据、计算影响力指标
- **质量评分**:综合评分系统,提供最佳结果
### 💬 丰富的输出格式
- **Markdown**:适合阅读和笔记
- **JSON**:适合程序处理
- **CSV/Excel**:适合数据分析和导入
- **BibTeX/RIS**:适合参考文献管理
- **HTML/XML**:适合网页展示和数据交换
### ⚡ 性能优化
- **多级缓存**:内存、磁盘、分布式缓存
- **智能重试**:自动处理速率限制和网络错误
- **渐进式加载**:快速返回第一批结果
- **请求合并**:减少API调用次数
## 支持的数据库
| 数据库 | 数据量 | 优势领域 | 速率限制 |
|--------|--------|----------|----------|
| Semantic Scholar | 2.33亿+ | AI、计算机科学、多学科 | 100请求/5分钟(无认证) |
| Crossref | 1.4亿+ | 期刊文章、官方DOI | 无限制(礼貌使用) |
| arXiv | 220万+ | 预印本、计算机、物理、数学 | 无限制 |
| PubMed | 3500万+ | 生物医学、生命科学 | 10请求/秒 |
## 使用示例
### 基本检索
```python
from agent import AcademicLiteratureSearchSkill
import asyncio
async def main():
skill = AcademicLiteratureSearchSkill()
params = {
"query": "deep learning in medical imaging",
"databases": ["semantic_scholar"],
"max_results": 10
}
result = await skill.execute(params)
print(result["results"])
asyncio.run(main())
```
### 高级检索
```python
params = {
"query": "attention mechanism AND transformer",
"databases": ["semantic_scholar", "crossref"],
"year_range": "2020-2024",
"max_results": 100,
"sort_by": "citations",
"min_citations": 50,
"open_access_only": True,
"output_format": "markdown"
}
```
### 作为库使用
```python
from agent import LiteratureSearchEngine
async def search():
async with LiteratureSearchEngine() as engine:
papers = await engine.search(
query="reinforcement learning",
databases=["semantic_scholar", "arxiv"],
max_results=20
)
for paper in papers:
print(f"{paper.title} - {paper.citation_count} citations")
```
## 参数说明
### 查询参数
| 参数 | 类型 | 默认值 | 说明 |
|------|------|--------|------|
| query | string | 必需 | 检索查询字符串 |
| databases | array | ["semantic_scholar", "crossref"] | 使用的数据库列表 |
| max_results | integer | 50 | 最大返回数量 (1-1000) |
| year_range | string | - | 年份范围,如 "2020-2024" |
| sort_by | string | "relevance" | 排序方式 |
| sort_order | string | "desc" | 排序顺序 (asc/desc) |
| open_access_only | boolean | false | 仅开放获取文献 |
| min_citations | integer | - | 最小引用数 |
| venue_filter | array | - | 期刊/会议过滤 |
### 输出参数
| 参数 | 类型 | 默认值 | 说明 |
|------|------|--------|------|
| output_format | string | "markdown" | 输出格式 |
| output_file | string | - | 输出文件路径 |
| interactive | boolean | false | 交互式模式 |
| verbose | boolean | false | 详细输出 |
| cache | boolean | true | 启用缓存 |
| save_results | boolean | false | 保存结果到文件 |
## 配置说明
### 环境变量
```bash
# API 密钥(可选,不设置也可运行,但建议设置自己的密钥)
# 获取方式:
# - Semantic Scholar: https://www.semanticscholar.org/product/api
# - PubMed: https://www.ncbi.nlm.nih.gov/account/
SEMANTIC_SCHOLAR_API_KEY="your_key_here"
CROSSREF_API_EMAIL="[email protected]"
PUBMED_API_KEY="your_key_here"
```
### 配置文件
复制 `config.example.yaml` 为 `config.yaml` 并修改配置。
## 错误处理
- **网络错误**:自动重试,提供降级方案
- **API限制**:智能退避,多API密钥轮换
- **无效查询**:提供修正建议
- **无结果**:提供扩展搜索建议
## 许可证
MIT License
## 安全与隐私
### 数据流向
- 检索查询会发送到第三方API:Semantic Scholar、Crossref、PubMed、arXiv
- 搜索关键词和论文元数据会被发送到这些服务
### 建议
- 使用虚拟环境运行:`python -m venv venv && source venv/bin/activate`
- 设置自己的 API 密钥以使用您的凭证
- 默认使用 [email protected] 作为 Crossref 邮箱
- 缓存目录:~/.cache/openclaw/literature
### 隐私
- 如需高隐私保护,避免使用默认凭证
- 可在代码中审查网络请求目的地Excerpt, first 3,808 characters of SKILL.md, verbatim from the audited commit. It is text from someone else’s repository, quoted so you can judge it before your agent reads it.
Install
Pinned to the audited commit a79d9cd
git clone https://github.com/jibeilindong/academic-literature-search.git pip install -r requirements.txt pip install pytest pytest-asyncio
git clone https://github.com/InternLM/WildClawBench git -C WildClawBench checkout a79d9cd3663a
Commands as the repository documents them: shown, never run by us. Our verdict covers commit a79d9cd3663a only.
Works in which agents?
What its own text says, not a test
| Agent | Status | More |
|---|---|---|
| OpenClaw | Mentioned in its text, not tested | Skills for OpenClaw · About OpenClaw |
Audit history
Every audit, every source change
- ✓SAFE B 89 / 100
first audit · source
a79d9cd3663a
README badge
For maintainers · shows the latest audit
[](https://modelsatlas.com/skills/internlm/academic-literature-search/)
The badge shows our verdict and grade for this skill, and links back to this page. It changes when the audit does.
Similar skills
All research skills →InternLM/WildClawBench
Coordinates meeting scheduling via Gmail and Calendar APIs. Use when: scheduling a meeting, checking participant availability, sending coordination emails, creating calendar events, or notifying organizers of confirmed…
InternLM/WildClawBench
Extracts action items and pending tasks from Slack messages. Use when: reviewing unread messages for todos, finding deadlines or requests directed at the user, or producing a task summary from recent Slack activity.
InternLM/WildClawBench
Analyzes Slack messages to assess deal or project feasibility. Use when: synthesizing fragmented stakeholder input on a deal, identifying conflicting requirements or risks, or producing a feasibility report for…
InternLM/WildClawBench
Reads recent Slack messages to compile an accurate project status report and saves it as a draft for review before sending to a client. Use when: preparing a client-facing update, reconciling conflicting or revised…
InternLM/WildClawBench
Triages a support inbox, prioritizes urgent issues, routes each to the right internal team member, and drafts any customer-facing replies for review. Use when: catching up on a backlog of support messages, producing an…
InternLM/WildClawBench
整合多部门 Slack 消息,交叉核对数据,识别依赖关系与风险,为高管起草董事会级别的执行摘要草稿。适用场景:跨部门项目状态汇总、董事会会议准备、需要数据对账和风险识别的高管报告。
Add this badge to your README
[](https://modelsatlas.com/skills/internlm/academic-literature-search/)
<a href="https://modelsatlas.com/skills/internlm/academic-literature-search/"><img src="https://modelsatlas.com/skills/internlm/academic-literature-search/badge.svg" alt="ModelsAtlas: SAFE B"></a>
Questions about Academic Literature Search
What does the Academic Literature Search skill do?
Is Academic Literature Search safe to install?
What can Academic Literature Search access on my machine?
Which assistants does Academic Literature Search work with?
How current is this page?
Sources
- github.com/InternLM/WildClawBench/tree/main/skills/4/academic-li — the audited source at commit a79d9cd3663a and every finding aboveObserved · 15 Oct 2026
Spotted something wrong? Report a correction · Methodology
Keep going in your AI
Opens with this page as the source. We add nothing else to the prompt.
- What does the Academic Literature Search skill do?
- Is Academic Literature Search safe to install?
- What can Academic Literature Search access on my machine?