multimodal-analysis
Analyzes medical images (X-ray, MRI, CT) using multimodal LLMs to identify anomalies and generate reports.
Lab Protocol from Literature - Extract and generate lab protocol: search PubMed, extract protocol from paper, and generate executable protocol. Use this skill for lab science tasks involving pubmed search extract protocol from pdf protocol generation generate executable json. Combines 4 tools fro...
---
name: lab-protocol-from-literature
description: "Lab Protocol from Literature - Extract and generate lab protocol: search PubMed, extract protocol from paper, and generate executable protocol. Use this skill for lab science tasks involving pubmed search extract protocol from pdf protocol generation generate executable json. Combines 4 tools from 2 SCP server(s)."
---
# Lab Protocol from Literature
**Discipline**: Lab Science | **Tools Used**: 4 | **Servers**: 2
## Description
Extract and generate lab protocol: search PubMed, extract protocol from paper, and generate executable protocol.
## Tools Used
- **`pubmed_search`** from `search-server` (streamable-http) - `https://scp.intern-ai.org.cn/api/v1/mcp/7/Origene-Search`
- **`extract_protocol_from_pdf`** from `server-19` (streamable-http) - `https://scp.intern-ai.org.cn/api/v1/mcp/19/Thoth-Plan`
- **`protocol_generation`** from `server-19` (streamable-http) - `https://scp.intern-ai.org.cn/api/v1/mcp/19/Thoth-Plan`
- **`generate_executable_json`** from `server-19` (streamable-http) - `https://scp.intern-ai.org.cn/api/v1/mcp/19/Thoth-Plan`
## Workflow
1. Search PubMed for protocols
2. Extract protocol from paper
3. Generate standard protocol
4. Create executable JSON
## Test Case
### Input
```json
{
"query": "PCR amplification protocol for BRCA1"
}
```
### Expected Steps
1. Search PubMed for protocols
2. Extract protocol from paper
3. Generate standard protocol
4. Create executable JSON
## Usage Example
> **Note:** Replace `<YOUR_SCP_HUB_API_KEY>` with your own SCP Hub API Key. You can obtain one from the [SCP Platform](https://scphub.intern-ai.org.cn).
```python
import asyncio
import json
from mcp import ClientSession
from mcp.client.streamable_http import streamablehttp_client
from mcp.client.sse import sse_client
SERVERS = {
"search-server": "https://scp.intern-ai.org.cn/api/v1/mcp/7/Origene-Search",
"server-19": "https://scp.intern-ai.org.cn/api/v1/mcp/19/Thoth-Plan"
}
async def connect(url, transport_type):
transport = streamablehttp_client(url=url, headers={"SCP-HUB-API-KEY": "<YOUR_SCP_HUB_API_KEY>"})
read, write, _ = await transport.__aenter__()
ctx = ClientSession(read, write)
session = await ctx.__aenter__()
await session.initialize()
return session, ctx, transport
def parse(result):
try:
if hasattr(result, 'content') and result.content:
c = result.content[0]
if hasattr(c, 'text'):
try: return json.loads(c.text)
except: return c.text
return str(result)
except: return str(result)
async def main():
# Connect to required servers
sessions = {}
sessions["search-server"], _, _ = await connect("https://scp.intern-ai.org.cn/api/v1/mcp/7/Origene-Search", "streamable-http")
sessions["server-19"], _, _ = await connect("https://scp.intern-ai.org.cn/api/v1/mcp/19/Thoth-Plan", "streamable-http")
# Execute workflow steps
# Step 1: Search PubMed for protocols
result_1 = await sessions["search-server"].call_tool("pubmed_search", arguments={})
data_1 = parse(result_1)
print(f"Step 1 result: {json.dumps(data_1, indent=2, ensure_ascii=False)[:500]}")
# Step 2: Extract protocol from paper
result_2 = await sessions["server-19"].call_tool("extract_protocol_from_pdf", arguments={})
data_2 = parse(result_2)
print(f"Step 2 result: {json.dumps(data_2, indent=2, ensure_ascii=False)[:500]}")
# Step 3: Generate standard protocol
result_3 = await sessions["server-19"].call_tool("protocol_generation", arguments={})
data_3 = parse(result_3)
print(f"Step 3 result: {json.dumps(data_3, indent=2, ensure_ascii=False)[:500]}")
# Step 4: Create executable JSON
result_4 = await sessions["server-19"].call_tool("generate_executable_json", arguments={})
data_4 = parse(result_4)
print(f"Step 4 result: {json.dumps(data_4, indent=2, ensure_ascii=False)[:500]}")
# Cleanup
print("Workflow complete!")
if __name__ == "__main__":
asyncio.run(main())
```
---
name: lab-protocol-from-literature
description: "Lab Protocol from Literature - Extract and generate lab protocol: search PubMed, extract protocol from paper, and generate executable protocol. Use this skill for lab science tasks involving pubmed search extract protocol from pdf protocol generation generate executable json. Combines 4 tools from 2 SCP server(s)."
---
# Lab Protocol from Literature
**Discipline**: Lab Science | **Tools Used**: 4 | **Servers**: 2
## Description
Extract and generate lab protocol: search PubMed, extract protocol from paper, and generate executable protocol.
## Tools Used
- **`pubmed_search`** from `search-server` (streamable-http) - `https://scp.intern-ai.org.cn/api/v1/mcp/7/Origene-Search`
- **`extract_protocol_from_pdf`** from `server-19` (streamable-http) - `https://scp.intern-ai.org.cn/api/v1/mcp/19/Thoth-Plan`
- **`protocol_generation`** from `server-19` (streamable-http) - `https://scp.intern-ai.org.cn/api/v1/mcp/19/Thoth-Plan`
- **`generate_executable_json`** from `server-19` (streamable-http) - `https://scp.intern-ai.org.cn/api/v1/mcp/19/Thoth-Plan`
## Workflow
1. Search PubMed for protocols
2. Extract protocol from paper
3. Generate standard protocol
4. Create executable JSON
## Test Case
### Input
```json
{
"query": "PCR amplification protocol for BRCA1"
}
```
### Expected Steps
1. Search PubMed for protocols
2. Extract protocol from paper
3. Generate standard protocol
4. Create executable JSON
## Usage Example
> **Note:** Replace `<YOUR_SCP_HUB_API_KEY>` with your own SCP Hub API Key. You can obtain one from the [SCP Platform](https://scphub.intern-ai.org.cn).
```python
import asyncio
import json
from mcp import ClientSession
from mcp.client.streamable_http import streamablehttp_client
from mcp.client.sse import sse_client
SERVERS = {
"search-server": "https://scp.intern-ai.org.cn/api/v1/mcp/7/Origene-Search",
"server-19": "https://scp.intern-ai.org.cn/api/v1/mcp/19/Thoth-Plan"
}
async def connect(url, transport_type):
transport = streamablehttp_client(url=url, headers={"SCP-HUB-API-KEY": "<YOUR_SCP_HUB_API_KEY>"})
read, write, _ = await transport.__aenter__()
ctx = ClientSession(read, write)
session = await ctx.__aenter__()
await session.initialize()
return session, ctx, transport
def parse(result):
try:
if hasattr(result, 'content') and result.content:
c = result.content[0]
if hasattr(c, 'text'):
try: return json.loads(c.text)
except: return c.text
return str(result)
except: return str(result)
async def main():
# Connect to required servers
sessions = {}
sessions["search-server"], _, _ = await connect("https://scp.intern-ai.org.cn/api/v1/mcp/7/Origene-Search", "streamable-http")
sessions["server-19"], _, _ = await connect("https://scp.intern-ai.org.cn/api/v1/mcp/19/Thoth-Plan", "streamable-http")
# Execute workflow steps
# Step 1: Search PubMed for protocols
result_1 = await sessions["search-server"].call_tool("pubmed_search", arguments={})
data_1 = parse(result_1)
print(f"Step 1 result: {json.dumps(data_1, indent=2, ensure_ascii=False)[:500]}")
# Step 2: Extract protocol from paper
result_2 = await sessions["server-19"].call_tool("extract_protocol_from_pdf", arguments={})
data_2 = parse(result_2)
print(f"Step 2 result: {json.dumps(data_2, indent=2, ensure_ascii=False)[:500]}")
# Step 3: Generate standard protocol
result_3 = await sessions["server-19"].call_tool("protocol_generation", arguments={})
data_3 = parse(result_3)
print(f"Step 3 result: {json.dumps(data_3, indent=2, ensure_ascii=False)[:500]}")
# Step 4: Create executable JSON
result_4 = await sessions["server-19"].call_tool("generate_executable_json", arguments={})
data_4 = parse(result_4)
print(f"Step 4 result: {json.dumps(data_4, indent=2, ensure_ascii=False)[:500]}")
# Cleanup
print("Workflow complete!")
if __name__ == "__main__":
asyncio.run(main())
```
Imported from the listed AgentSkill.sh category. Complete supplied package files and notices are retained. Required external tools and credentials are not installed. No bundled code was executed during import. Individual agent task behavior has not been tested.