Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

MDR 2017/745 Specialist · borghei | DecimalAI

Legalby borghei

MDR 2017/745 Specialist

EU MDR 2017/745 compliance specialist for medical device classification, technical documentation, clinical evidence, and post-market surveillance.

Never made a case worseMaintained

SkillScore

SkillScore64/100

Test results 64

Safety

3/3 safety checks passed

Code scan
Intent
Content

What it changes

55%64%of test cases passed +9 pts

Without skill12 of 22

With skill14 of 22

2 fixed12 passed either way8 still failing0 broke

Measured on gemini-3.6-flash8/20/2026v1 · latest is v2

This benchmark measured an earlier version of the skill.

Track record

Agent runs: 0; All time

See one task, with and without the skill

One test taskWhat is the official European Commission deadline date when the first four operational EUDAMED modules (Actor, UDI/Device, Certificates, Clinical Investigations) become mandatory for economic operators across the EU?

Without the skillMissed

With the skillPassed

Same model, same task. Answers are capped and redacted.

Use this skill

Your agent runs on

Adds every file the skill has, pinned so you can update it. No account needed.

pip install "decimalai>=0.15"
decimalai skills add borghei-mdr-745-specialist
claude --plugin-dir "$(git rev-parse --show-toplevel 2>/dev/null || pwd)/.agents"

The same install from Python, for a notebook or a setup script.

Install

pip install "decimalai>=0.15"

Add the skill

from decimalai import skillpkg

# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("borghei-mdr-745-specialist")
if result.exit_code:
    raise SystemExit(result.problems)

print("Added. Start Claude Code with the command below.")

Start Claude Code with it

claude --plugin-dir "$(git rev-parse --show-toplevel 2>/dev/null || pwd)/.agents"

Type these in a Claude Code session. No account or API key. It writes the skill and any files it bundles, and asks first when its safety rating isn't Passed.

Add the DecimalAI plugin, once

/plugin marketplace add decimal-labs/decimalai-claude-plugins
/plugin install decimalai@decimalai

Install this skill

/decimalai:install borghei-mdr-745-specialist

The plugin needs python3 and curl on your machine.

No Python or account. Claude Code picks it up in its next session.

Save it into your project

curl -fsSL --create-dirs -o .claude/skills/borghei-mdr-745-specialist/SKILL.md https://app.decimal.ai/s/borghei-mdr-745-specialist/SKILL.md

Copy SKILL.md

Save it in your project, as

.claude/skills/borghei-mdr-745-specialist/SKILL.md

This gets SKILL.md only. The skill's 4 other files are not included.

Other ways to add it

Claude Code docs

agent_config() hands your agent the skill. With no DecimalAI key it reads your project's files and sends nothing. Needs SDK 0.27 or newer.

Install

pip install "decimalai[claude-agent-sdk]"
decimalai skills add borghei-mdr-745-specialist

Add it to your agent

import asyncio

from claude_agent_sdk import ClaudeAgentOptions
from decimalai.claude_agent_sdk import agent_config

# With no DECIMAL_API_KEY this reads the skills your project lists, from disk.
support = agent_config("support", fallback="You are a helpful agent.")

async def main():
    question = "Hello!"
    async with support.run(question) as run:  # one user message, one run
        async for message in run.query(options=ClaudeAgentOptions(max_turns=6)):
            print(message)

# In a notebook or async code: await main()
asyncio.run(main())

Prefer its own loader? ClaudeAgentOptions(plugins=[skills_plugin()], skills="all") reads the same folder. Use one or the other: a skill delivered both ways reaches the prompt twice.

The same install from Python, for a notebook or a setup script. agent_config() hands your agent the skill. With no DecimalAI key it reads your project's files and sends nothing. Needs SDK 0.27 or newer.

Install

pip install "decimalai[claude-agent-sdk]"

Add it to your agent

from decimalai import skillpkg

# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("borghei-mdr-745-specialist")
if result.exit_code:
    raise SystemExit(result.problems)

import asyncio

from claude_agent_sdk import ClaudeAgentOptions
from decimalai.claude_agent_sdk import agent_config

# With no DECIMAL_API_KEY this reads the skills your project lists, from disk.
support = agent_config("support", fallback="You are a helpful agent.")

async def main():
    question = "Hello!"
    async with support.run(question) as run:  # one user message, one run
        async for message in run.query(options=ClaudeAgentOptions(max_turns=6)):
            print(message)

# In a notebook or async code: await main()
asyncio.run(main())

Prefer its own loader? ClaudeAgentOptions(plugins=[skills_plugin()], skills="all") reads the same folder. Use one or the other: a skill delivered both ways reaches the prompt twice.

No DecimalAI package. The whole skill goes in every prompt, and nothing pins or updates it.

Copy SKILL.md

Save it in your project, as

.agents/skills/borghei-mdr-745-specialist/SKILL.md

Give it to your agent

import asyncio
from pathlib import Path
from claude_agent_sdk import ClaudeAgentOptions, query

skill = Path(".agents/skills/borghei-mdr-745-specialist/SKILL.md").read_text()
options = ClaudeAgentOptions(system_prompt=f"You are a helpful agent.\n\n{skill}", max_turns=6)

async def main():
    async for message in query(prompt="Hello!", options=options):
        print(message)

# In a notebook or async code: await main()
asyncio.run(main())

This gets SKILL.md only. The skill's 4 other files are not included.

Other ways to add it

Claude Agent SDK docs

agent_config() hands your agent the skill. With no DecimalAI key it reads your project's files and sends nothing. Needs SDK 0.27 or newer.

Install

pip install "decimalai[openai-agents]"
decimalai skills add borghei-mdr-745-specialist

Add it to your agent

from agents import Agent, Runner
from decimalai.openai_agents import agent_config

# With no DECIMAL_API_KEY this reads the skills your project lists, from disk.
support = agent_config("support", fallback="You are a helpful agent.")

agent = Agent(
    name=support.name,
    instructions=support.instructions,  # your prompt, then the skill list
    tools=[*support.tools],             # your tools, then load_skill
)
# In a notebook or async code: result = await Runner.run(agent, "Hello!")
result = Runner.run_sync(agent, "Hello!")

Using SandboxAgent? capabilities=[Shell(), decimalai.openai_agents.sandbox_skills()] points its own loader at the same folder.

Install

pip install "decimalai[openai-agents]"

Add it to your agent

from decimalai import skillpkg

# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("borghei-mdr-745-specialist")
if result.exit_code:
    raise SystemExit(result.problems)

from agents import Agent, Runner
from decimalai.openai_agents import agent_config

# With no DECIMAL_API_KEY this reads the skills your project lists, from disk.
support = agent_config("support", fallback="You are a helpful agent.")

agent = Agent(
    name=support.name,
    instructions=support.instructions,  # your prompt, then the skill list
    tools=[*support.tools],             # your tools, then load_skill
)
# In a notebook or async code: result = await Runner.run(agent, "Hello!")
result = Runner.run_sync(agent, "Hello!")

Using SandboxAgent? capabilities=[Shell(), decimalai.openai_agents.sandbox_skills()] points its own loader at the same folder.

No DecimalAI package. The whole skill goes in every prompt, and nothing pins or updates it.

Copy SKILL.md

Save it in your project, as

.agents/skills/borghei-mdr-745-specialist/SKILL.md

Give it to your agent

from pathlib import Path
from agents import Agent, Runner

skill = Path(".agents/skills/borghei-mdr-745-specialist/SKILL.md").read_text()
agent = Agent(name="support", instructions=f"You are a helpful agent.\n\n{skill}")
# In a notebook or async code: result = await Runner.run(agent, "Hello!")
result = Runner.run_sync(agent, "Hello!")

This gets SKILL.md only. The skill's 4 other files are not included.

Other ways to add it

OpenAI Agents docs

agent_config() hands your agent the skill. With no DecimalAI key it reads your project's files and sends nothing. Needs SDK 0.27 or newer.

Install

pip install "decimalai[openai]"
decimalai skills add borghei-mdr-745-specialist

Add it to your agent

import json

from openai import OpenAI
from decimalai.openai import agent_config

client = OpenAI()
# With no DECIMAL_API_KEY this reads the skills your project lists, from disk.
support = agent_config("support", fallback="You are a helpful agent.")


def lookup_order(order_id: str) -> str:
    return f"Order {order_id} shipped yesterday."


# Your tools, as Chat Completions takes them, and what answers each.
TOOLS = [{"type": "function", "function": {
    "name": "lookup_order",
    "description": "Look up the status of an order.",
    "parameters": {"type": "object", "properties": {"order_id": {"type": "string"}},
                   "required": ["order_id"]},
}}]
FUNCTIONS = {"lookup_order": lookup_order}

question = "Where is order A1001?"
with support.run(question) as run:  # one user message, one run
    # run.system is your prompt, then the skill list.
    messages = [{"role": "system", "content": run.system}, run.message]
    while True:
        reply = client.chat.completions.create(
            model="gpt-5.4-mini",
            messages=messages,
            tools=[*TOOLS, *run.tools],  # your tools, then load_skill
        ).choices[0].message
        messages.append(reply)
        if not reply.tool_calls:
            break
        for call in reply.tool_calls:
            # load_skill is answered by the run; your own tools by FUNCTIONS.
            function = FUNCTIONS.get(call.function.name)
            messages.append(run.tool_result(call) or {
                "role": "tool", "tool_call_id": call.id,
                "content": function(**json.loads(call.function.arguments)) if function
                else f"There is no tool named {call.function.name!r}.",
            })
print(reply.content)

# The Responses API takes run.response_tools instead of run.tools.

Install

pip install "decimalai[openai]"

Add it to your agent

from decimalai import skillpkg

# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("borghei-mdr-745-specialist")
if result.exit_code:
    raise SystemExit(result.problems)

import json

from openai import OpenAI
from decimalai.openai import agent_config

client = OpenAI()
# With no DECIMAL_API_KEY this reads the skills your project lists, from disk.
support = agent_config("support", fallback="You are a helpful agent.")


def lookup_order(order_id: str) -> str:
    return f"Order {order_id} shipped yesterday."


# Your tools, as Chat Completions takes them, and what answers each.
TOOLS = [{"type": "function", "function": {
    "name": "lookup_order",
    "description": "Look up the status of an order.",
    "parameters": {"type": "object", "properties": {"order_id": {"type": "string"}},
                   "required": ["order_id"]},
}}]
FUNCTIONS = {"lookup_order": lookup_order}

question = "Where is order A1001?"
with support.run(question) as run:  # one user message, one run
    # run.system is your prompt, then the skill list.
    messages = [{"role": "system", "content": run.system}, run.message]
    while True:
        reply = client.chat.completions.create(
            model="gpt-5.4-mini",
            messages=messages,
            tools=[*TOOLS, *run.tools],  # your tools, then load_skill
        ).choices[0].message
        messages.append(reply)
        if not reply.tool_calls:
            break
        for call in reply.tool_calls:
            # load_skill is answered by the run; your own tools by FUNCTIONS.
            function = FUNCTIONS.get(call.function.name)
            messages.append(run.tool_result(call) or {
                "role": "tool", "tool_call_id": call.id,
                "content": function(**json.loads(call.function.arguments)) if function
                else f"There is no tool named {call.function.name!r}.",
            })
print(reply.content)

# The Responses API takes run.response_tools instead of run.tools.

No DecimalAI package. The whole skill goes in every prompt, and nothing pins or updates it.

Copy SKILL.md

Save it in your project, as

.agents/skills/borghei-mdr-745-specialist/SKILL.md

Give it to your agent

from pathlib import Path
from openai import OpenAI

client = OpenAI()
skill = Path(".agents/skills/borghei-mdr-745-specialist/SKILL.md").read_text()
reply = client.chat.completions.create(
    model="gpt-5.4-mini",
    messages=[
        {"role": "system", "content": f"You are a helpful agent.\n\n{skill}"},
        {"role": "user", "content": "Hello!"},
    ],
).choices[0].message

This gets SKILL.md only. The skill's 4 other files are not included.

Other ways to add it

OpenAI SDK docs

agent_config() hands your agent the skill. With no DecimalAI key it reads your project's files and sends nothing. Needs SDK 0.27 or newer.

Install

pip install "decimalai[anthropic]"
decimalai skills add borghei-mdr-745-specialist

Add it to your agent

import anthropic
from decimalai.anthropic import agent_config

client = anthropic.Anthropic()
# With no DECIMAL_API_KEY this reads the skills your project lists, from disk.
support = agent_config("support", fallback="You are a helpful agent.")

question = "Hello!"
with support.run(question) as run:  # one user message, one run
    answer = client.beta.messages.tool_runner(
        model="claude-haiku-5-5",
        max_tokens=4096,
        system=run.system,     # your prompt, then the skill list
        tools=[*run.tools],    # your tools, then load_skill
        messages=[{"role": "user", "content": question}],
    ).until_done()

Install

pip install "decimalai[anthropic]"

Add it to your agent

from decimalai import skillpkg

# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("borghei-mdr-745-specialist")
if result.exit_code:
    raise SystemExit(result.problems)

import anthropic
from decimalai.anthropic import agent_config

client = anthropic.Anthropic()
# With no DECIMAL_API_KEY this reads the skills your project lists, from disk.
support = agent_config("support", fallback="You are a helpful agent.")

question = "Hello!"
with support.run(question) as run:  # one user message, one run
    answer = client.beta.messages.tool_runner(
        model="claude-haiku-5-5",
        max_tokens=4096,
        system=run.system,     # your prompt, then the skill list
        tools=[*run.tools],    # your tools, then load_skill
        messages=[{"role": "user", "content": question}],
    ).until_done()

No DecimalAI package. The whole skill goes in every prompt, and nothing pins or updates it.

Copy SKILL.md

Save it in your project, as

.agents/skills/borghei-mdr-745-specialist/SKILL.md

Give it to your agent

from pathlib import Path
import anthropic

client = anthropic.Anthropic()
skill = Path(".agents/skills/borghei-mdr-745-specialist/SKILL.md").read_text()
reply = client.messages.create(
    model="claude-haiku-5-5",
    max_tokens=4096,
    system=f"You are a helpful agent.\n\n{skill}",
    messages=[{"role": "user", "content": "Hello!"}],
)

This gets SKILL.md only. The skill's 4 other files are not included.

Other ways to add it

Anthropic docs

agent_config() hands your agent the skill. With no DecimalAI key it reads your project's files and sends nothing. Needs SDK 0.27 or newer.

Install

pip install "decimalai[google-genai]"
decimalai skills add borghei-mdr-745-specialist

Add it to your agent

from google import genai
from google.genai import types
from decimalai.google_genai import agent_config

client = genai.Client()
# With no DECIMAL_API_KEY this reads the skills your project lists, from disk.
support = agent_config("support", fallback="You are a helpful agent.")


def lookup_order(order_id: str) -> str:
    return f"Order {order_id} shipped yesterday."


# Your functions, as declarations for the model, and what answers each.
TOOLS = [{"function_declarations": [{
    "name": "lookup_order",
    "description": "Look up the status of an order.",
    "parameters_json_schema": {"type": "object", "properties": {"order_id": {"type": "string"}},
                               "required": ["order_id"]},
}]}]
FUNCTIONS = {"lookup_order": lookup_order}

question = "Where is order A1001?"
with support.run(question) as run:  # one user message, one run
    config = types.GenerateContentConfig(
        system_instruction=run.system,     # your prompt, then the skill list
        tools=[*TOOLS, *run.tool_params],  # your functions, then load_skill
    )
    contents = [{"role": "user", "parts": [{"text": run.message}]}]
    while True:
        response = client.models.generate_content(
            model="gemini-3.5-flash", contents=contents, config=config,
        )
        if not response.function_calls:
            break
        contents.append(response.candidates[0].content)
        parts = []
        for call in response.function_calls:
            # load_skill is answered by the run; your own functions by FUNCTIONS.
            # A loop of your own rather than automatic function calling, which
            # raises KeyError when the model asks for a function you did not pass.
            function = FUNCTIONS.get(call.name)
            parts.append(run.tool_result(call) or {"function_response": {
                "name": call.name, "id": call.id,
                "response": {"result": function(**(call.args or {})) if function
                             else f"There is no function named {call.name!r}."},
            }})
        contents.append({"role": "user", "parts": parts})
print(response.text)

Install

pip install "decimalai[google-genai]"

Add it to your agent

from decimalai import skillpkg

# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("borghei-mdr-745-specialist")
if result.exit_code:
    raise SystemExit(result.problems)

from google import genai
from google.genai import types
from decimalai.google_genai import agent_config

client = genai.Client()
# With no DECIMAL_API_KEY this reads the skills your project lists, from disk.
support = agent_config("support", fallback="You are a helpful agent.")


def lookup_order(order_id: str) -> str:
    return f"Order {order_id} shipped yesterday."


# Your functions, as declarations for the model, and what answers each.
TOOLS = [{"function_declarations": [{
    "name": "lookup_order",
    "description": "Look up the status of an order.",
    "parameters_json_schema": {"type": "object", "properties": {"order_id": {"type": "string"}},
                               "required": ["order_id"]},
}]}]
FUNCTIONS = {"lookup_order": lookup_order}

question = "Where is order A1001?"
with support.run(question) as run:  # one user message, one run
    config = types.GenerateContentConfig(
        system_instruction=run.system,     # your prompt, then the skill list
        tools=[*TOOLS, *run.tool_params],  # your functions, then load_skill
    )
    contents = [{"role": "user", "parts": [{"text": run.message}]}]
    while True:
        response = client.models.generate_content(
            model="gemini-3.5-flash", contents=contents, config=config,
        )
        if not response.function_calls:
            break
        contents.append(response.candidates[0].content)
        parts = []
        for call in response.function_calls:
            # load_skill is answered by the run; your own functions by FUNCTIONS.
            # A loop of your own rather than automatic function calling, which
            # raises KeyError when the model asks for a function you did not pass.
            function = FUNCTIONS.get(call.name)
            parts.append(run.tool_result(call) or {"function_response": {
                "name": call.name, "id": call.id,
                "response": {"result": function(**(call.args or {})) if function
                             else f"There is no function named {call.name!r}."},
            }})
        contents.append({"role": "user", "parts": parts})
print(response.text)

No DecimalAI package. The whole skill goes in every prompt, and nothing pins or updates it.

Copy SKILL.md

Save it in your project, as

.agents/skills/borghei-mdr-745-specialist/SKILL.md

Give it to your agent

from pathlib import Path
from google import genai
from google.genai import types

client = genai.Client()
skill = Path(".agents/skills/borghei-mdr-745-specialist/SKILL.md").read_text()
response = client.models.generate_content(
    model="gemini-3.5-flash",
    contents="Hello!",
    config=types.GenerateContentConfig(system_instruction=f"You are a helpful agent.\n\n{skill}"),
)

This gets SKILL.md only. The skill's 4 other files are not included.

Other ways to add it

Gemini API docs

agent_config() hands your agent the skill. With no DecimalAI key it reads your project's files and sends nothing. Needs SDK 0.27 or newer.

Install

pip install "decimalai[langchain]" langchain-openai
decimalai skills add borghei-mdr-745-specialist

Add it to your agent

from langchain.agents import create_agent
from decimalai.langchain import agent_config

# With no DECIMAL_API_KEY this reads the skills your project lists, from disk.
support = agent_config("support", fallback="You are a helpful agent.")

agent = create_agent(
    "openai:gpt-5.4-mini",
    tools=[],                         # your tools; the middleware adds load_skill
    middleware=[support.middleware],  # your prompt, the skill list and load_skill
)
result = agent.invoke({"messages": [{"role": "user", "content": "Hello!"}]})

Install

pip install "decimalai[langchain]" langchain-openai

Add it to your agent

from decimalai import skillpkg

# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("borghei-mdr-745-specialist")
if result.exit_code:
    raise SystemExit(result.problems)

from langchain.agents import create_agent
from decimalai.langchain import agent_config

# With no DECIMAL_API_KEY this reads the skills your project lists, from disk.
support = agent_config("support", fallback="You are a helpful agent.")

agent = create_agent(
    "openai:gpt-5.4-mini",
    tools=[],                         # your tools; the middleware adds load_skill
    middleware=[support.middleware],  # your prompt, the skill list and load_skill
)
result = agent.invoke({"messages": [{"role": "user", "content": "Hello!"}]})

No DecimalAI package. The whole skill goes in every prompt, and nothing pins or updates it.

Copy SKILL.md

Save it in your project, as

.agents/skills/borghei-mdr-745-specialist/SKILL.md

Give it to your agent

from pathlib import Path
from langchain.agents import create_agent

skill = Path(".agents/skills/borghei-mdr-745-specialist/SKILL.md").read_text()
agent = create_agent("openai:gpt-5.4-mini", tools=[], system_prompt=f"You are a helpful agent.\n\n{skill}")
result = agent.invoke({"messages": [{"role": "user", "content": "Hello!"}]})

This gets SKILL.md only. The skill's 4 other files are not included.

Other ways to add it

LangChain docs

CrewAI's own skill loader reads the installed folder. No DecimalAI account needed.

Install

pip install "decimalai>=0.16" "crewai>=1.15.27"
decimalai skills add borghei-mdr-745-specialist

Add it to your agent

import decimalai.crewai
from crewai import Agent, Crew, Task

agent = Agent(
    role="Support", goal="Help the customer", backstory="A support agent.",
    llm=llm,                               # your own CrewAI LLM
    skills=decimalai.crewai.skills(),      # the installed folder
    tools=decimalai.crewai.skill_tools(),  # reads a skill's files, runs its scripts
)
task = Task(description="Hello!", expected_output="an answer", agent=agent)
Crew(agents=[agent], tasks=[task]).kickoff()

The same install from Python, for a notebook or a setup script. CrewAI's own skill loader reads the installed folder. No DecimalAI account needed.

Install

pip install "decimalai>=0.16" "crewai>=1.15.27"

Add it to your agent

from decimalai import skillpkg

# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("borghei-mdr-745-specialist")
if result.exit_code:
    raise SystemExit(result.problems)

import decimalai.crewai
from crewai import Agent, Crew, Task

agent = Agent(
    role="Support", goal="Help the customer", backstory="A support agent.",
    llm=llm,                               # your own CrewAI LLM
    skills=decimalai.crewai.skills(),      # the installed folder
    tools=decimalai.crewai.skill_tools(),  # reads a skill's files, runs its scripts
)
task = Task(description="Hello!", expected_output="an answer", agent=agent)
Crew(agents=[agent], tasks=[task]).kickoff()

No DecimalAI package. CrewAI's own loader reads the folder, and nothing pins or updates it.

Copy SKILL.md

Save it in your project, as

.agents/skills/borghei-mdr-745-specialist/SKILL.md

Give it to your agent

from pathlib import Path
from crewai import Agent, Crew, Task

agent = Agent(
    role="Support", goal="Help the customer", backstory="A support agent.",
    llm=llm,                         # your own CrewAI LLM
    skills=[Path(".agents/skills")],
)
task = Task(description="Hello!", expected_output="an answer", agent=agent)
Crew(agents=[agent], tasks=[task]).kickoff()

This gets SKILL.md only. The skill's 4 other files are not included.

Other ways to add it

CrewAI docs

agent_config() hands your agent the skill. With no DecimalAI key it reads your project's files and sends nothing. Needs SDK 0.27 or newer.

Install

pip install "decimalai[adk]"
decimalai skills add borghei-mdr-745-specialist

Add it to your agent

from decimalai.adk import DecimalAIPlugin, agent_config
from google.adk.agents import LlmAgent
from google.adk.apps import App

# With no DECIMAL_API_KEY this reads the skills your project lists, from disk.
support = agent_config("support", fallback="You are a helpful agent.")

agent = LlmAgent(
    name="support",          # ADK needs a Python identifier
    model="gemini-3.5-flash",
    instruction=support.instruction,  # your prompt, then the skill list
    tools=[*support.tools],           # your tools, then load_skill
)
# DecimalAIPlugin makes each invocation of runners built from this App one trace.
app = App(name="support", root_agent=agent, plugins=[DecimalAIPlugin()])

Prefer ADK's own loader? SkillToolset(skills=decimalai.adk.skills(), ...) reads the same folder. Use one or the other: a skill delivered both ways reaches the prompt twice.

Install

pip install "decimalai[adk]"

Add it to your agent

from decimalai import skillpkg

# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("borghei-mdr-745-specialist")
if result.exit_code:
    raise SystemExit(result.problems)

from decimalai.adk import DecimalAIPlugin, agent_config
from google.adk.agents import LlmAgent
from google.adk.apps import App

# With no DECIMAL_API_KEY this reads the skills your project lists, from disk.
support = agent_config("support", fallback="You are a helpful agent.")

agent = LlmAgent(
    name="support",          # ADK needs a Python identifier
    model="gemini-3.5-flash",
    instruction=support.instruction,  # your prompt, then the skill list
    tools=[*support.tools],           # your tools, then load_skill
)
# DecimalAIPlugin makes each invocation of runners built from this App one trace.
app = App(name="support", root_agent=agent, plugins=[DecimalAIPlugin()])

Prefer ADK's own loader? SkillToolset(skills=decimalai.adk.skills(), ...) reads the same folder. Use one or the other: a skill delivered both ways reaches the prompt twice.

No DecimalAI package. The whole skill goes in every prompt, and nothing pins or updates it.

Copy SKILL.md

Save it in your project, as

.agents/skills/borghei-mdr-745-specialist/SKILL.md

Give it to your agent

from pathlib import Path
from google.adk.agents import LlmAgent

skill = Path(".agents/skills/borghei-mdr-745-specialist/SKILL.md").read_text()
agent = LlmAgent(
    name="support",
    model="gemini-3.5-flash",
    instruction=lambda context: f"You are a helpful agent.\n\n{skill}",
)

This gets SKILL.md only. The skill's 4 other files are not included.

Other ways to add it

Google ADK docs

agent_config() hands your agent the skill. With no DecimalAI key it reads your project's files and sends nothing. Needs SDK 0.27 or newer.

Install

pip install "decimalai[pydantic-ai]"
decimalai skills add borghei-mdr-745-specialist

Add it to your agent

from pydantic_ai import Agent
from decimalai.pydantic_ai import agent_config

# With no DECIMAL_API_KEY this reads the skills your project lists, from disk.
support = agent_config("support", fallback="You are a helpful agent.")

agent = Agent(
    "openai:gpt-4o-mini",
    name=support.name,
    tools=[],                 # your tools; the capability adds load_skill
    capabilities=[support],   # your prompt, the skill list and load_skill
)
# In a notebook or async code: result = await agent.run("Hello!")
result = agent.run_sync("Hello!")

On the Pydantic AI harness, decimalai.pydantic_ai_harness.skills(project) uses its own loader instead.

Install

pip install "decimalai[pydantic-ai]"

Add it to your agent

from decimalai import skillpkg

# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("borghei-mdr-745-specialist")
if result.exit_code:
    raise SystemExit(result.problems)

from pydantic_ai import Agent
from decimalai.pydantic_ai import agent_config

# With no DECIMAL_API_KEY this reads the skills your project lists, from disk.
support = agent_config("support", fallback="You are a helpful agent.")

agent = Agent(
    "openai:gpt-4o-mini",
    name=support.name,
    tools=[],                 # your tools; the capability adds load_skill
    capabilities=[support],   # your prompt, the skill list and load_skill
)
# In a notebook or async code: result = await agent.run("Hello!")
result = agent.run_sync("Hello!")

On the Pydantic AI harness, decimalai.pydantic_ai_harness.skills(project) uses its own loader instead.

No DecimalAI package. The whole skill goes in every prompt, and nothing pins or updates it.

Copy SKILL.md

Save it in your project, as

.agents/skills/borghei-mdr-745-specialist/SKILL.md

Give it to your agent

from pathlib import Path
from pydantic_ai import Agent

skill = Path(".agents/skills/borghei-mdr-745-specialist/SKILL.md").read_text()
agent = Agent("openai:gpt-4o-mini", instructions=f"You are a helpful agent.\n\n{skill}")
# In a notebook or async code: result = await agent.run("Hello!")
result = agent.run_sync("Hello!")

This gets SKILL.md only. The skill's 4 other files are not included.

Other ways to add it

Pydantic AI docs

Deep Agents' own skill loader reads the installed folder. No DecimalAI account needed.

Install

pip install "decimalai>=0.16" deepagents
decimalai skills add borghei-mdr-745-specialist

Add it to your agent

import decimalai.deepagents
from deepagents import create_deep_agent

agent = create_deep_agent(model=model, **decimalai.deepagents.skills())
agent.invoke({"messages": [{"role": "user", "content": "Hello!"}]})

The default backend reads files but has no shell. To run a skill's scripts, pass skills(backend=LocalShellBackend(...)).

The same install from Python, for a notebook or a setup script. Deep Agents' own skill loader reads the installed folder. No DecimalAI account needed.

Install

pip install "decimalai>=0.16" deepagents

Add it to your agent

from decimalai import skillpkg

# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("borghei-mdr-745-specialist")
if result.exit_code:
    raise SystemExit(result.problems)

import decimalai.deepagents
from deepagents import create_deep_agent

agent = create_deep_agent(model=model, **decimalai.deepagents.skills())
agent.invoke({"messages": [{"role": "user", "content": "Hello!"}]})

The default backend reads files but has no shell. To run a skill's scripts, pass skills(backend=LocalShellBackend(...)).

No DecimalAI package. The whole skill goes in every prompt, and nothing pins or updates it.

Copy SKILL.md

Save it in your project, as

.agents/skills/borghei-mdr-745-specialist/SKILL.md

Give it to your agent

from pathlib import Path
from deepagents import create_deep_agent

skill = Path(".agents/skills/borghei-mdr-745-specialist/SKILL.md").read_text()
agent = create_deep_agent(model=model, system_prompt=f"You are a helpful agent.\n\n{skill}")
agent.invoke({"messages": [{"role": "user", "content": "Hello!"}]})

This gets SKILL.md only. The skill's 4 other files are not included.

Other ways to add it

Deep Agents docs

Its own skill loader reads the installed folder. No DecimalAI account needed.

Install

pip install "decimalai>=0.16" agent-framework
decimalai skills add borghei-mdr-745-specialist

Add it to your agent

import asyncio

from agent_framework import Agent
from decimalai.agent_framework import skills

agent = Agent(client=client, instructions="You are a helpful agent.", context_providers=[skills()])
asyncio.run(agent.run("Hello!"))

It caches skills by default, so restart the process after decimalai skills update.

The same install from Python, for a notebook or a setup script. Its own skill loader reads the installed folder. No DecimalAI account needed.

Install

pip install "decimalai>=0.16" agent-framework

Add it to your agent

from decimalai import skillpkg

# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("borghei-mdr-745-specialist")
if result.exit_code:
    raise SystemExit(result.problems)

import asyncio

from agent_framework import Agent
from decimalai.agent_framework import skills

agent = Agent(client=client, instructions="You are a helpful agent.", context_providers=[skills()])
asyncio.run(agent.run("Hello!"))

It caches skills by default, so restart the process after decimalai skills update.

No DecimalAI package. The whole skill goes in every prompt, and nothing pins or updates it.

Copy SKILL.md

Save it in your project, as

.agents/skills/borghei-mdr-745-specialist/SKILL.md

Give it to your agent

import asyncio
from pathlib import Path
from agent_framework import Agent

skill = Path(".agents/skills/borghei-mdr-745-specialist/SKILL.md").read_text()
agent = Agent(client=client, instructions=f"You are a helpful agent.\n\n{skill}")
asyncio.run(agent.run("Hello!"))

This gets SKILL.md only. The skill's 4 other files are not included.

Other ways to add it

Microsoft Agent Framework docs

Strands' own skill loader reads the installed folder. No DecimalAI account needed.

Install

pip install "decimalai>=0.16" strands-agents
decimalai skills add borghei-mdr-745-specialist

Add it to your agent

from decimalai.strands import skills
from strands import Agent
from strands.vended_tools import file_editor, shell

agent = Agent(
    model=model,                # always pass one: the default is Amazon Bedrock
    system_prompt="You are a helpful agent.",
    tools=[file_editor, shell],
    plugins=[skills()],
)
agent("Hello!")

Strands loads an agent's skills once, so build the agent again after decimalai skills update.

The same install from Python, for a notebook or a setup script. Strands' own skill loader reads the installed folder. No DecimalAI account needed.

Install

pip install "decimalai>=0.16" strands-agents

Add it to your agent

from decimalai import skillpkg

# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("borghei-mdr-745-specialist")
if result.exit_code:
    raise SystemExit(result.problems)

from decimalai.strands import skills
from strands import Agent
from strands.vended_tools import file_editor, shell

agent = Agent(
    model=model,                # always pass one: the default is Amazon Bedrock
    system_prompt="You are a helpful agent.",
    tools=[file_editor, shell],
    plugins=[skills()],
)
agent("Hello!")

Strands loads an agent's skills once, so build the agent again after decimalai skills update.

No DecimalAI package. The whole skill goes in every prompt, and nothing pins or updates it.

Copy SKILL.md

Save it in your project, as

.agents/skills/borghei-mdr-745-specialist/SKILL.md

Give it to your agent

from pathlib import Path
from strands import Agent

skill = Path(".agents/skills/borghei-mdr-745-specialist/SKILL.md").read_text()
agent = Agent(model=model, system_prompt=f"You are a helpful agent.\n\n{skill}")
agent("Hello!")

This gets SKILL.md only. The skill's 4 other files are not included.

Other ways to add it

Strands docs

Adds every file the skill has, pinned so you can update it. Cursor, Copilot, Cline and Warp read this folder by themselves.

pip install "decimalai>=0.15"
decimalai skills add borghei-mdr-745-specialist

No Python or account. Cursor, Copilot, Cline and Warp read this folder by themselves.

Save it into your project

curl -fsSL --create-dirs -o .agents/skills/borghei-mdr-745-specialist/SKILL.md https://app.decimal.ai/s/borghei-mdr-745-specialist/SKILL.md

Copy SKILL.md

Save it in your project, as

.agents/skills/borghei-mdr-745-specialist/SKILL.md

This gets SKILL.md only. The skill's 4 other files are not included.

Other ways to add it

Cursor and other editors docs

Fetch the file and give its text to your model as instructions. No account or package needed.

Get the file

curl -fsSL https://app.decimal.ai/s/borghei-mdr-745-specialist/SKILL.md

Read the summary as JSON

curl -fsSL https://app.decimal.ai/s/borghei-mdr-745-specialist.json

This gets SKILL.md only. The skill's 4 other files are not included.

Other ways to add it

Other setups docs

You don't need a DecimalAI account.

Version: v2 (v1 tested 8/20/2026)
Source: borghei/claude-skills ↗
License: NOASSERTION

Manage it in DecimalAI

Saved prompts and skills, pinned versions, and traces of this skill in your runs. DecimalAI does not run your agent.

Claude Code reads files from your project, so DecimalAI config does not apply there.

On Claude Agent SDK: save the skill to an agent here, set DECIMAL_API_KEY, and the same code uses it with no command to run. Changes you save reach the next run with no redeploy.

On OpenAI Agents: save the skill to an agent here, set DECIMAL_API_KEY, and the same code uses it with no command to run. Changes you save reach the next run with no redeploy.

On OpenAI SDK: save the skill to an agent here, set DECIMAL_API_KEY, and the same code uses it with no command to run. Changes you save reach the next run with no redeploy.

On Anthropic: save the skill to an agent here, set DECIMAL_API_KEY, and the same code uses it with no command to run. Changes you save reach the next run with no redeploy.

On Gemini API: save the skill to an agent here, set DECIMAL_API_KEY, and the same code uses it with no command to run. Changes you save reach the next run with no redeploy.

On LangChain: save the skill to an agent here, set DECIMAL_API_KEY, and the same code uses it with no command to run. Changes you save reach the next run with no redeploy.

On CrewAI: set DECIMAL_API_KEY and add decimalai.init(crewai=True) to record runs and skill use. The skill version stays in your repo.

On Google ADK: save the skill to an agent here, set DECIMAL_API_KEY, and the same code uses it with no command to run. Changes you save reach the next run with no redeploy.

On Pydantic AI: save the skill to an agent here, set DECIMAL_API_KEY, and the same code uses it with no command to run. Changes you save reach the next run with no redeploy.

On Deep Agents: set DECIMAL_API_KEY and add decimalai.init(langchain=True) to record runs and skill use. The skill version stays in your repo.

On Microsoft Agent Framework: set DECIMAL_API_KEY and add decimalai.init(otel=True, agent_name="support") to record runs and skill use. The skill version stays in your repo.

On Strands: set DECIMAL_API_KEY and add decimalai.init(otel=True, agent_name="support") to record runs and skill use. The skill version stays in your repo.

Editors read files from your project, so DecimalAI config does not apply there.

Anywhere else: record a use with decimalai.log_skill_activation() inside a trace.

Use this skill

SKILL.md

MDR 2017/745 Specialist

EU MDR compliance patterns for medical device classification, technical documentation, and clinical evidence.

Device Classification Workflow
Technical Documentation
Clinical Evidence
Post-Market Surveillance
EUDAMED and UDI
Reference Documentation
Tools

Clarify First

Before classifying the device or analyzing gaps, settle the inputs below. When one is missing or vague and would change the result, ask about it; when it wouldn't, use a sensible default and say which one you used.

Device characteristics: duration, invasiveness, body-system contact, and whether it is active/software (drives the Annex VIII classification).
Intended purpose: the device's claimed clinical purpose (sets the applicable rule and clinical-evidence level).
Software/AI nature: whether it is software as a medical device (MDCG 2019-11) or AI/ML (changes classification and the documentation set).

Ask about the two or three inputs that matter most for this request, not the whole list. If the user wants a result straight away, produce it and state your assumptions at the top.

Device Classification Workflow

Classify device under MDR Annex VIII:

Identify device duration (transient, short-term, long-term)
Determine invasiveness level (non-invasive, body orifice, surgical)
Assess body system contact (CNS, cardiac, other)
Check if active device (energy dependent)
Apply classification rules 1-22
For software, apply MDCG 2019-11 algorithm
Document classification rationale
Validation: Classification confirmed with Notified Body

Classification Matrix

Factor	Class I	Class IIa	Class IIb	Class III
Duration	Any	Short-term	Long-term	Long-term
Invasiveness	Non-invasive	Body orifice	Surgical	Implantable
System	Any	Non-critical	Critical organs	CNS/cardiac
Risk	Lowest	Low-medium	Medium-high	Highest

Software Classification (MDCG 2019-11)

Information Use	Condition Severity	Class
Informs decision	Non-serious	IIa
Informs decision	Serious	IIb
Drives/treats	Critical	III

Classification Examples

Example 1: Absorbable Surgical Suture

Rule 8 (implantable, long-term)
Duration: > 30 days (absorbed)
Contact: General tissue
Classification: Class IIb

Example 2: AI Diagnostic Software

Rule 11 + MDCG 2019-11
Function: Diagnoses serious condition
Classification: Class IIb

Example 3: Cardiac Pacemaker

Rule 8 (implantable)
Contact: Central circulatory system
Classification: Class III

Technical Documentation

Prepare technical file per Annex II and III:

Create device description (variants, accessories, intended purpose)
Develop labeling (Article 13 requirements, IFU)
Document design and manufacturing process
Complete GSPR compliance matrix
Prepare benefit-risk analysis
Compile verification and validation evidence
Integrate risk management file (ISO 14971)
Validation: Technical file reviewed for completeness

Technical File Structure

ANNEX II TECHNICAL DOCUMENTATION
├── Device description and UDI-DI
├── Label and instructions for use
├── Design and manufacturing info
├── GSPR compliance matrix
├── Benefit-risk analysis
├── Verification and validation
└── Clinical evaluation report

GSPR Compliance Checklist

Requirement	Evidence	Status
Safe design (GSPR 1-3)	Risk management file	☐
Chemical properties (GSPR 10.1)	Biocompatibility report	☐
Infection risk (GSPR 10.2)	Sterilization validation	☐
Software requirements (GSPR 17)	IEC 62304 documentation	☐
Labeling (GSPR 23)	Label artwork, IFU	☐

Conformity Assessment Routes

Class	Route	NB Involvement
I	Annex II self-declaration	None
Is/Im	Annex II + IX/XI	Sterile/measuring aspects
IIa	Annex II + IX or XI	Product or QMS
IIb	Annex IX + X or X + XI	Type exam + production
III	Annex IX + X	Full QMS + type exam

Clinical Evidence

Develop clinical evidence strategy per Annex XIV:

Define clinical claims and endpoints
Conduct systematic literature search
Appraise clinical data quality
Assess equivalence (technical, biological, clinical)
Identify evidence gaps
Determine if clinical investigation required
Prepare Clinical Evaluation Report (CER)
Validation: CER reviewed by qualified evaluator

Evidence Requirements by Class

Class	Minimum Evidence	Investigation
I	Risk-benefit analysis	Not typically required
IIa	Literature + post-market	May be required
IIb	Systematic literature review	Often required
III	Comprehensive clinical data	Required (Article 61)

Clinical Evaluation Report Structure

CER CONTENTS
├── Executive summary
├── Device scope and intended purpose
├── Clinical background (state of the art)
├── Literature search methodology
├── Data appraisal and analysis
├── Safety and performance conclusions
├── Benefit-risk determination
└── PMCF plan summary

Qualified Evaluator Requirements

Medical degree or equivalent healthcare qualification
4+ years clinical experience in relevant field
Training in clinical evaluation methodology
Understanding of MDR requirements

Post-Market Surveillance

Establish PMS system per Chapter VII:

Develop PMS plan (Article 84)
Define data collection methods
Establish complaint handling procedures
Create vigilance reporting process
Plan Periodic Safety Update Reports (PSUR)
Integrate with PMCF activities
Define trend analysis and signal detection
Validation: PMS system audited annually

PMS System Components

Component	Requirement	Frequency
PMS Plan	Article 84	Maintain current
PSUR	Class IIa and higher	Per class schedule
PMCF Plan	Annex XIV Part B	Update with CER
PMCF Report	Annex XIV Part B	Annual (Class III)
Vigilance	Articles 87-92	As events occur

PSUR Schedule

Class	Frequency
Class III	Annual
Class IIb implantable	Annual
Class IIb	Every 2 years
Class IIa	When necessary

Serious Incident Reporting

Timeline	Requirement
2 days	Serious public health threat
10 days	Death or serious deterioration
15 days	Other serious incidents

EUDAMED and UDI

Implement UDI system per Article 27:

Obtain issuing entity code (GS1, HIBCC, ICCBBA)
Assign UDI-DI to each device variant
Assign UDI-PI (production identifier)
Apply UDI carrier to labels (AIDC + HRI)
Register actor in EUDAMED
Register devices in EUDAMED
Upload certificates when available
Validation: UDI verified on sample labels

EUDAMED Modules

Module	Content	Actor
Actor	Company registration	Manufacturer, AR
UDI/Device	Device and variant data	Manufacturer
Certificates	NB certificates	Notified Body
Clinical Investigation	Study registration	Sponsor
Vigilance	Incident reports	Manufacturer
Market Surveillance	Authority actions	Competent Authority

UDI Label Requirements

Required elements per Article 13:

UDI-DI (device identifier)
UDI-PI (production identifier) for Class II+
AIDC format (barcode/RFID)
HRI format (human-readable)
Manufacturer name and address
Lot/serial number
Expiration date (if applicable)

Reference Documentation

MDR Classification Guide

references/mdr-classification-guide.md contains:

Complete Annex VIII classification rules (Rules 1-22)
Software classification per MDCG 2019-11
Worked classification examples
Conformity assessment route selection

Clinical Evidence Requirements

references/clinical-evidence-requirements.md contains:

Clinical evidence framework and hierarchy
Literature search methodology
Clinical Evaluation Report structure
PMCF plan and evaluation report guidance

Technical Documentation Templates

references/technical-documentation-templates.md contains:

Annex II and III content requirements
Design History File structure
GSPR compliance matrix template
Declaration of Conformity template
Notified Body submission checklist

Tools

MDR Gap Analyzer

# Quick gap analysis
python scripts/mdr_gap_analyzer.py --device "Device Name" --class IIa

# JSON output for integration
python scripts/mdr_gap_analyzer.py --device "Device Name" --class III --output json

# Interactive assessment
python scripts/mdr_gap_analyzer.py --interactive

Analyzes device against MDR requirements, identifies compliance gaps, generates prioritized recommendations.

Output includes:

Requirements checklist by category
Gap identification with priorities
Critical gap highlighting
Compliance roadmap recommendations

Notified Body Interface

Selection Criteria

Factor	Considerations
Designation scope	Covers your device type
Capacity	Timeline for initial audit
Geographic reach	Markets you need to access
Technical expertise	Experience with your technology
Fee structure	Transparency, predictability

Pre-Submission Checklist

Technical documentation complete
GSPR matrix fully addressed
Risk management file current
Clinical evaluation report complete
QMS (ISO 13485) certified
Labeling and IFU finalized
Validation: Internal gap assessment complete

MDCG Guidance Documents Update (2024-2025)

Key MDCG Guidance Documents

Document	Title	Status	Key Impact
MDCG 2024-1	Transition provisions under MDR Art. 120	Final (2024)	Extended transition deadlines for legacy devices
MDCG 2024-6	Clinical evidence needed for medical devices previously CE marked under Directives	Final (2024)	Reduced clinical evidence burden for well-established devices
MDCG 2023-4 Rev.1	Notified Body capacity and availability	Revised (2024)	NB capacity monitoring and optimization
MDCG 2022-18 Rev.1	Software qualification and classification under MDR	Revised (2024)	Updated software classification algorithm
MDCG 2020-1 Rev.1	Clinical evaluation — equivalence	Revised (2024)	Refined equivalence demonstration requirements
MDCG 2019-16 Rev.1	Cybersecurity for medical devices	Revised (2024)	Enhanced cybersecurity requirements for connected devices
MDCG 2019-11 Rev.1	Qualification and classification of software	Active	Software as medical device classification

MDR Transition Timeline (Post-Amendment Regulation 2023/607)

Device Category	Transition Deadline	Conditions
Class III and implantable	26 May 2026	Valid MDD/AIMDD certificate + QMS application to NB by 26 May 2024
Class IIb	31 December 2027	Valid certificate + QMS application to NB
Class IIa and Class I (sterile/measuring)	31 December 2028	Valid certificate + QMS application to NB
Class I (up-classified under MDR)	31 December 2028	Previously exempt from NB involvement

Conditions for extended deadlines:

Device continues to comply with MDD/AIMDD
No significant changes in design or intended purpose
No unacceptable safety or performance risk
Manufacturer has applied to Notified Body for MDR conformity assessment before applicable deadline

Software as Medical Device Under MDR (MDCG 2019-11 Rev.1)

Software Qualification Decision

Is the software a medical device?
        │
        ▼
Does the software perform an action on data?
        │
    Yes─┴─No → NOT a medical device (data storage/communication only)
     │
     ▼
Is the action for the benefit of individual patients?
        │
    Yes─┴─No → NOT a medical device (population health/admin)
     │
     ▼
Is the action one of: treatment, diagnosis, monitoring, prediction?
        │
    Yes─┴─No → NOT a medical device (lifestyle/wellness)
     │
     ▼
QUALIFIES AS MEDICAL DEVICE SOFTWARE → Apply classification rules

Software Classification Under MDR

Decision Factor	Class IIa	Class IIb	Class III
Provides information to inform clinical management	Non-serious conditions	Serious conditions	N/A
Drives clinical management or diagnoses	N/A	Non-serious conditions	Serious or critical conditions
Monitors physiological processes	Non-vital parameters	Vital parameters (not immediate danger)	Vital parameters (immediate danger)

Software Lifecycle Requirements Under MDR

MDR Requirement	IEC 62304 Mapping	Documentation
GSPR 17.1 (repeatability, reliability)	Software development process	Software development plan
GSPR 17.2 (state of the art)	Software architecture	Architecture design document
GSPR 17.3 (minimum IT requirements)	System requirements	IT environment specification
GSPR 17.4 (foreseeable misuse)	Risk management	Software risk analysis
Annex II §6.5 (software verification)	Software testing	Test plans and reports
Annex I §23.4 (labeling for software)	Release documentation	Software release notes

AI/ML Medical Devices Under MDR

AI/ML Classification Considerations

AI/ML Capability	MDR Classification Impact	Regulatory Consideration
AI-assisted detection	Typically Class IIa-IIb (Rule 11)	Must demonstrate clinical performance per intended use
AI-driven diagnosis	Typically Class IIb-III (Rule 11)	Requires clinical investigation for novel indications
AI treatment optimization	Typically Class IIb-III (Rule 11 + specific rules)	Benefit-risk analysis must account for AI uncertainty
Continuously learning AI	Classification per highest risk output	Post-market monitoring must track algorithm evolution

AI/ML-Specific Technical Documentation

In addition to standard Annex II requirements, AI/ML devices must document:

Section	Content	MDCG Reference
Algorithm description	Architecture, training approach, feature engineering	MDCG 2019-11, IMDRF SaMD WG
Training data	Sources, demographics, size, labeling methodology, quality	MDCG 2020-1 (equivalence)
Validation methodology	Test dataset independence, performance metrics, subgroup analysis	Annex XIV
Clinical performance	Sensitivity, specificity, AUC, PPV, NPV per intended population	CER requirements
Change management	How algorithm updates are validated and deployed	GSPR 17
Explainability	How the AI's output can be understood by intended users	GSPR 23 (labeling)

EU AI Act Interaction with MDR

EU AI Act Requirement	MDR Equivalent	Combined Approach
High-risk classification (Annex III, Point 10)	Annex VIII classification rules	Both classifications apply; meet stricter requirement
Conformity assessment (Art. 43)	Annex IX/X/XI assessment	MDR conformity assessment satisfies AI Act (Art. 120)
Technical documentation (Annex IV)	Annex II technical documentation	Extend MDR technical file with AI Act-specific elements
Risk management (Art. 9)	ISO 14971 + GSPR	ISO 14971 satisfies both when AI risks are included
Data governance (Art. 10)	GSPR 17 + Annex XIV	Add AI-specific data governance to clinical evaluation
Post-market monitoring (Art. 72)	Chapter VII PMS	Single PMS system covering both AI Act and MDR

See also: ../risk-management-specialist/SKILL.md for AI-specific risk management under ISO 14971, and ../fda-consultant-specialist/SKILL.md for FDA's AI/ML SaMD framework and PCCP.

Eudamed Implementation Status and Requirements

Eudamed Module Deployment Status (as of 2025)

Module	Status	Mandatory Date	Content
Actor Registration	Operational	Available now	Economic operator registration
UDI/Device Registration	Operational	6 months after Eudamed fully functional	Device and UDI-DI data
Notified Body and Certificates	Operational	Available now	Certificate data upload by NBs
Clinical Investigations	Operational	Available now	Study registration and reporting
Vigilance	Partially operational	24 months after fully functional	Incident reports, FSCAs, trend reports
Market Surveillance	In development	18 months after fully functional	CA market surveillance activities

Key consideration: Until Eudamed is declared fully functional by the European Commission, manufacturers must use existing national systems (e.g., BfArM in Germany, ANSM in France) for vigilance reporting.

Eudamed Registration Requirements for Manufacturers

Data Element	Required For	Update Frequency
SRN (Single Registration Number)	All manufacturers	On change
Authorized representative details	Non-EU manufacturers	On change
Device identification (UDI-DI)	All devices placed on market	Before first placing on market
Basic UDI-DI	Device model/family grouping	Before first placing on market
GMDN code	Device nomenclature	On initial registration
Risk class	Classification per Annex VIII	On initial registration
NB certificate reference	Class IIa and above	When certificate issued
Clinical investigation registration	Interventional studies	Before study start

UDI-DI and UDI-PI Detailed Requirements

UDI-DI (Device Identifier) — Static Information

Element	Description	Example
Device identifier	Unique code for device model/version	04069876543219 (GS1 GTIN)
Issuing entity	GS1, HIBCC, ICCBBA, or IFA	Selected by manufacturer
Device model	Specific device configuration	"CardioMonitor X200"
Device version	Software version (for SaMD)	"v3.1.0"
Applicable regulations	MDR or IVDR reference	MDR 2017/745
Risk class	Per Annex VIII	Class IIa
Basic UDI-DI	Grouping identifier for device family	04069876500001

UDI-PI (Production Identifier) — Variable Information

PI Element	When Required	Format
Lot/batch number	When tracking by lot	Lot: ABC123
Serial number	When individual tracking required (Class III, implantable)	SN: 2024-00001
Manufacturing date	When relevant to safety	Mfg: 2024-06-15
Expiry date	When device has shelf life	Exp: 2026-06-15
Software version	SaMD and software-driven devices	SW: v3.1.0

UDI Carrier Requirements

Carrier Type	Format	Where Applied
AIDC (barcode/2D code)	GS1 DataMatrix, GS1-128, HIBC	Device label, package label
HRI (human-readable)	Plain text interpretation of AIDC	Adjacent to AIDC on label
RFID	GS1 EPC/RFID	Optional, in addition to AIDC

Labeling placement rules:

UDI on each level of packaging (unit, intermediate, case)
For reusable devices requiring sterilization: UDI on device itself (direct marking)
Class III implantable: UDI on device or packaging that remains with patient record
UDI must survive device lifecycle (including reprocessing cycles for reusable devices)

UDI Database Submission Timeline

Device Class	Submission Deadline
Class III and implantable	Before placing on the market
Class IIb	Before placing on the market
Class IIa	Before placing on the market
Class I	Before placing on the market

Note: All timelines are contingent on Eudamed being declared fully functional. Until then, manufacturers should pre-register in Eudamed (available modules) and maintain data readiness.

Cross-Reference: NIS2 for Critical Infrastructure (Healthcare)

Healthcare organizations manufacturing or deploying medical devices may be classified as essential entities under NIS2:

NIS2 Requirement	MDR Impact	Action for Manufacturers
Art. 21: Cybersecurity risk management	MDCG 2019-16 cybersecurity guidance	Align device cybersecurity with organizational NIS2 compliance
Art. 23: Incident reporting (24h/72h)	Art. 87-92 vigilance reporting	Unified incident reporting covering both device and infrastructure incidents
Art. 21(2)(d): Supply chain security	Art. 11 authorized representatives, supply chain	Assess cybersecurity of device component suppliers
Art. 20: Governance and accountability	Art. 10 manufacturer obligations	Senior management oversight of both NIS2 and MDR compliance

See also: ../information-security-manager-iso27001/SKILL.md for ISO 27001 alignment with NIS2 requirements.

MDR Updates & Cross-Framework Integration

Latest MDCG Guidance Documents

MDCG 2019-11 Rev.1: Qualification and classification of software — updated algorithm for SaMD
MDCG 2020-1 Rev.1: Clinical evaluation (Annex XIV) — updated methodologies
MDCG 2024-8: EU AI Act interaction with MDR for AI-enabled medical devices
MDCG 2023-4: Legacy devices and Article 120 transition provisions

AI/ML Medical Devices Under MDR

Classification: AI/ML-based SaMD typically Class IIa or higher (Rule 11)
Clinical Evidence: Must demonstrate AI algorithm clinical performance (sensitivity, specificity, AUC)
Continuous Learning: PCCP-equivalent under MDR for adaptive AI devices
Post-Market: Enhanced PMCF for AI devices monitoring real-world algorithm performance
EU AI Act Interaction: High-risk AI medical devices subject to BOTH MDR and EU AI Act
Cross-reference: See eu-ai-act-specialist for AI Act obligations

NIS2 Impact on Healthcare

Healthcare entities are "Essential Entities" under NIS2 Directive
Medical device manufacturers may be "Important Entities"
NIS2 cybersecurity requirements supplement MDR cybersecurity expectations
Cross-reference: See nis2-directive-specialist for NIS2 compliance

EUDAMED Implementation Status

Actor Registration Module: Operational — all economic operators must register
UDI/Device Registration Module: Operational — mandatory device registration
Notified Body Module: Operational — certificate information
Clinical Investigation Module: Available
Vigilance Module: Under development
Market Surveillance Module: Under development

UDI Detailed Requirements

UDI-DI (Device Identifier): Unique to device model — used for EUDAMED registration
UDI-PI (Production Identifier): Identifies production unit — lot, serial, expiry, manufacturing date
Issuing Entities: GS1, HIBCC, ICCBBA, IFA
Carrier Types: AIDC (barcode/2D) + HRI (human readable)
Implant Card: Required for Class III implantable devices (UDI + patient information)

Troubleshooting

Problem	Possible Cause	Resolution
Device classification unclear -- rules yield different results	Multiple classification rules apply; highest class must be selected per MDR Article 51(7)	Apply all applicable rules from Annex VIII (Rules 1-22); use the implementing rule that gives the highest classification; for software, apply MDCG 2019-11 Rev.1 algorithm; document rationale for each rule considered
Notified Body rejects technical file for incompleteness	GSPR compliance matrix gaps, missing clinical evaluation, or insufficient risk management documentation	Review GSPR checklist in this skill; ensure Annex II technical file structure is complete; verify CER meets Annex XIV requirements; confirm ISO 14971 risk management file is current and comprehensive
EUDAMED registration delays blocking market access	Module not operational or manufacturer SRN not obtained	Check current EUDAMED module status; obtain SRN via actor registration module (operational); use national systems for vigilance reporting until Eudamed Vigilance module is fully functional
Clinical evidence insufficient for Class IIb/III device	Equivalence route rejected by NB or clinical investigation not planned	Reassess equivalence per MDCG 2020-1 Rev.1 (technical, biological, clinical equivalence with access to data); if equivalence fails, plan clinical investigation per Article 61; consider MDCG 2024-6 for reduced evidence burden on well-established devices
MDR transition deadline approaching with NB application pending	Limited NB capacity (approximately 40 designated EU-wide as of 2025)	Verify transition deadline for your device class (Class III: May 2026, Class IIb: Dec 2027, Class IIa: Dec 2028); ensure QMS application was submitted to NB by applicable deadline; maintain MDD/AIMDD compliance during transition
UDI labeling rejected by NB or competent authority	UDI-DI/UDI-PI format incorrect, AIDC carrier unreadable, or missing elements	Verify all required UDI elements per Article 27; ensure AIDC format (GS1 DataMatrix preferred) is scannable; include HRI adjacent to barcode; for reusable devices, apply direct marking that survives reprocessing
AI/ML medical device faces dual regulatory obligations	Device classified under both MDR and EU AI Act as high-risk	Assess both MDR Annex VIII classification and EU AI Act Annex III categorization; MDR conformity assessment may satisfy AI Act per Art. 120; extend technical documentation with AI-specific elements per MDCG 2024-8

Success Criteria

Device correctly classified with documented rationale -- classification per Annex VIII with all applicable rules evaluated, highest class selected, and rationale documented for NB review
Complete technical file per Annex II structure -- device description, labeling/IFU, design and manufacturing info, GSPR compliance matrix, benefit-risk analysis, verification and validation, and clinical evaluation report all present and current
GSPR compliance matrix fully addressed -- all applicable General Safety and Performance Requirements mapped to evidence with cross-references to risk management file, biocompatibility reports, sterilization validation, software documentation, and labeling
Clinical evaluation report meets Annex XIV requirements -- literature search methodology documented, data appraised and analyzed, safety and performance conclusions stated, benefit-risk determined, and PMCF plan included
PMS system operational -- PMS plan per Article 84, complaint handling procedures, vigilance reporting process, PSUR schedule defined by class, and PMCF activities integrated with CER
UDI system fully implemented -- UDI-DI assigned per device variant, UDI-PI applied (lot/serial/dates), AIDC and HRI carriers on all packaging levels, EUDAMED registration complete (when applicable)
MDR gap analysis shows zero critical gaps -- as measured by mdr_gap_analyzer.py, with all requirements addressed or in-progress with documented timeline

Scope & Limitations

In Scope:

Device classification per MDR Annex VIII (Rules 1-22) including software classification per MDCG 2019-11 Rev.1
Technical documentation structure and requirements per Annex II and Annex III
GSPR compliance matrix with evidence mapping
Clinical evidence strategy including equivalence assessment, CER structure, and PMCF planning
Post-market surveillance system design including PMS plan, PSUR schedule, and vigilance reporting timelines
EUDAMED and UDI system implementation guidance
Conformity assessment route selection by device class
MDR transition timeline tracking (post-Amendment Regulation 2023/607)
AI/ML medical device considerations including EU AI Act interaction

Out of Scope:

Clinical investigation protocol design, execution, or statistical analysis
Biocompatibility testing per ISO 10993 (beyond evidence mapping in GSPR)
Sterilization validation per ISO 11135/11137 (beyond evidence mapping)
Notified Body selection, engagement, or commercial negotiation
Quality Management System implementation -- use quality-manager-qms-iso13485 for ISO 13485 QMS
Risk management process implementation -- use risk-management-specialist for ISO 14971

Important Notes:

EUDAMED's first four modules became mandatory from May 28, 2026; manufacturers must have SRN and device registrations ready
Only approximately 40 Notified Bodies are designated EU-wide as of 2025, creating capacity constraints; early NB engagement is critical
The European Commission published updated transition timelines in December 2025 extending deadlines for certain device categories
Manufacturers must adopt recently harmonized standards with no formal transition period (Decision EU 2025/2078)

Integration Points

Skill	Integration	When to Use
`fda-consultant-specialist`	Cross-framework mapping for dual US/EU market; FDA QMSR aligns with ISO 13485 used by MDR	When device requires both FDA clearance/approval and EU MDR CE marking
`quality-manager-qms-iso13485`	ISO 13485 QMS is prerequisite for MDR conformity assessment (Annex IX, XI)	When establishing or auditing QMS for MDR compliance
`risk-management-specialist`	ISO 14971 risk management file is core component of MDR technical documentation	When developing risk management file, FMEA, and benefit-risk analysis
`eu-ai-act-specialist`	AI medical devices subject to both MDR and EU AI Act; classification and conformity assessment interaction	When AI-enabled medical device requires dual regulatory compliance
`capa-officer`	CAPA process supports MDR vigilance obligations and FSCA implementation	When post-market surveillance identifies safety or performance issues requiring corrective action
`infrastructure-compliance-auditor`	Cybersecurity validation per MDCG 2019-16 Rev.1 for connected medical devices	When connected device requires cybersecurity documentation for technical file

Tool Reference

mdr_gap_analyzer.py

Analyzes device against MDR requirements, identifies compliance gaps, and generates prioritized recommendations.

Flag	Required	Description
`--device <name>`	Yes (unless `--interactive`)	Device name for gap analysis
`--class <class>`	Yes (unless `--interactive`)	Device classification: `I`, `Is`, `Im`, `IIa`, `IIb`, `III`
`--output <format>`	No	Output format: `json` for machine-readable output; default is human-readable text
`--interactive`	No	Launch interactive assessment mode with guided questions

Analysis Categories: Technical documentation (Annex II), GSPR compliance, clinical evidence (Annex XIV), post-market surveillance (Chapter VII), UDI/EUDAMED, labeling (Article 13), quality management system, risk management, and conformity assessment route.

Output: Requirements checklist with per-item status (Not Started/In Progress/Complete/N/A), gap identification with priority (Critical/High/Medium/Low), critical gap highlighting, completion percentage, and compliance roadmap recommendations.

When it activates

Automatically, when a task matches. Use for Annex VIII classification, GSPR, clinical evaluation, EUDAMED, and UDI.

About this skill

Automatic

Clinical / Healthcare pack

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v8 — 10/25/2026

View 15 scanner checks

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

borghei / claude-skills

ra-qm-team/mdr-745-specialist/SKILL.md

View on GitHub ↗

License MIT + Commons ClauseImported 10/25/2026Version v2Bundled 4 files · scripts

Illustrative Want to see what borghei/mdr-745-specialist would put in front of your model? Preview the rendered body and check its trigger phrases — no LLM call, no signup.

Other skills for the same job

Published skills whose descriptions closely match this one, with their headline benchmark lift.

Related skills

Skills that do similar work, with their headline benchmark lift.

From the same source and nearby in the catalogue.

Ready to supercharge your AI agents?

SkillScore

Safety

What it changes

Track record

Use this skill

MDR 2017/745 Specialist

Table of Contents

Clarify First

Device Classification Workflow

Classification Matrix

Software Classification (MDCG 2019-11)

Classification Examples

Technical Documentation

Technical File Structure

GSPR Compliance Checklist

Conformity Assessment Routes

Clinical Evidence

Evidence Requirements by Class

Clinical Evaluation Report Structure

Qualified Evaluator Requirements

Post-Market Surveillance

PMS System Components

PSUR Schedule

Serious Incident Reporting

EUDAMED and UDI

EUDAMED Modules

UDI Label Requirements

Reference Documentation

MDR Classification Guide

Clinical Evidence Requirements

Technical Documentation Templates

Tools

MDR Gap Analyzer

Notified Body Interface

Selection Criteria

Pre-Submission Checklist

MDCG Guidance Documents Update (2024-2025)

Key MDCG Guidance Documents

MDR Transition Timeline (Post-Amendment Regulation 2023/607)

Software as Medical Device Under MDR (MDCG 2019-11 Rev.1)

Software Qualification Decision

Software Classification Under MDR

Software Lifecycle Requirements Under MDR

AI/ML Medical Devices Under MDR

AI/ML Classification Considerations

AI/ML-Specific Technical Documentation

EU AI Act Interaction with MDR

Eudamed Implementation Status and Requirements

Eudamed Module Deployment Status (as of 2025)

Eudamed Registration Requirements for Manufacturers

UDI-DI and UDI-PI Detailed Requirements

UDI-DI (Device Identifier) — Static Information

UDI-PI (Production Identifier) — Variable Information

UDI Carrier Requirements

UDI Database Submission Timeline

Cross-Reference: NIS2 for Critical Infrastructure (Healthcare)

MDR Updates & Cross-Framework Integration

Latest MDCG Guidance Documents

AI/ML Medical Devices Under MDR

NIS2 Impact on Healthcare

EUDAMED Implementation Status

UDI Detailed Requirements

Troubleshooting

Success Criteria

Scope & Limitations

Integration Points

Tool Reference

mdr_gap_analyzer.py

When it activates

SkillSafety

Source

Other skills for the same job

Related skills

More in the registry