Ready to supercharge your AI agents?

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

Material Design 3 · sentenz | DecimalAI

Frontendby sentenz

Material Design 3

Implement Google's Material Design 3 (Material You) UI system. Primary: Jetpack Compose Material3 (MaterialTheme, components, adaptive layout). Also Flutter and limited web (@material/web, maintenance mode). Covers tokens, 30+ components, layout, theming, M3 Expressive (platform matrix), and accessibility.

Maintained

SkillScore

SkillScore77/100

Test results 77

Safety

3/3 safety checks passed

Code scan
Intent
Content

What it changes

64%77%of test cases passed +14 pts

Without skill14 of 22

With skill17 of 22

5 fixed12 passed either way3 still failing2 broke

Measured on gemini-3.6-flash9/27/2026v121 comparable

Track record

Agent runs: 0; All time

See one task, with and without the skill

One test taskWe are planning to build refreshed search app bar layouts and animated top search views for a new multi-platform strategy. Which framework implementation target receives primary updates for expressive search and search app bars?

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 sentenz-material-3
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("sentenz-material-3")
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 sentenz-material-3

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/sentenz-material-3/SKILL.md https://app.decimal.ai/s/sentenz-material-3/SKILL.md

Copy SKILL.md

Save it in your project, as

.claude/skills/sentenz-material-3/SKILL.md

This gets SKILL.md only. The skill's 6 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 sentenz-material-3

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("sentenz-material-3")
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/sentenz-material-3/SKILL.md

Give it to your agent

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

skill = Path(".agents/skills/sentenz-material-3/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 6 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 sentenz-material-3

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("sentenz-material-3")
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/sentenz-material-3/SKILL.md

Give it to your agent

from pathlib import Path
from agents import Agent, Runner

skill = Path(".agents/skills/sentenz-material-3/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 6 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 sentenz-material-3

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("sentenz-material-3")
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/sentenz-material-3/SKILL.md

Give it to your agent

from pathlib import Path
from openai import OpenAI

client = OpenAI()
skill = Path(".agents/skills/sentenz-material-3/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 6 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 sentenz-material-3

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("sentenz-material-3")
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/sentenz-material-3/SKILL.md

Give it to your agent

from pathlib import Path
import anthropic

client = anthropic.Anthropic()
skill = Path(".agents/skills/sentenz-material-3/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 6 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 sentenz-material-3

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("sentenz-material-3")
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/sentenz-material-3/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/sentenz-material-3/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 6 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 sentenz-material-3

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("sentenz-material-3")
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/sentenz-material-3/SKILL.md

Give it to your agent

from pathlib import Path
from langchain.agents import create_agent

skill = Path(".agents/skills/sentenz-material-3/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 6 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 sentenz-material-3

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("sentenz-material-3")
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/sentenz-material-3/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 6 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 sentenz-material-3

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("sentenz-material-3")
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/sentenz-material-3/SKILL.md

Give it to your agent

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

skill = Path(".agents/skills/sentenz-material-3/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 6 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 sentenz-material-3

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("sentenz-material-3")
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/sentenz-material-3/SKILL.md

Give it to your agent

from pathlib import Path
from pydantic_ai import Agent

skill = Path(".agents/skills/sentenz-material-3/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 6 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 sentenz-material-3

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("sentenz-material-3")
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/sentenz-material-3/SKILL.md

Give it to your agent

from pathlib import Path
from deepagents import create_deep_agent

skill = Path(".agents/skills/sentenz-material-3/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 6 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 sentenz-material-3

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("sentenz-material-3")
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/sentenz-material-3/SKILL.md

Give it to your agent

import asyncio
from pathlib import Path
from agent_framework import Agent

skill = Path(".agents/skills/sentenz-material-3/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 6 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 sentenz-material-3

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("sentenz-material-3")
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/sentenz-material-3/SKILL.md

Give it to your agent

from pathlib import Path
from strands import Agent

skill = Path(".agents/skills/sentenz-material-3/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 6 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 sentenz-material-3

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/sentenz-material-3/SKILL.md https://app.decimal.ai/s/sentenz-material-3/SKILL.md

Copy SKILL.md

Save it in your project, as

.agents/skills/sentenz-material-3/SKILL.md

This gets SKILL.md only. The skill's 6 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/sentenz-material-3/SKILL.md

Read the summary as JSON

curl -fsSL https://app.decimal.ai/s/sentenz-material-3.json

This gets SKILL.md only. The skill's 6 other files are not included.

Other ways to add it

Other setups docs

You don't need a DecimalAI account.

Version: v1, tested 9/27/2026
Source: sentenz/skills ↗
License: Apache-2.0

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

Material Design 3

This skill guides implementation of Google's Material Design 3 (MD3) — a personal, adaptive, expressive design system. MD3 uses dynamic color, tonal surfaces, rounded shapes, and spring-based motion to create UIs that feel alive and personal.

Philosophy

MD3 is built on three principles:

Personal: Dynamic color adapts UI to the user's wallpaper or content. Theming is individual, not one-size-fits-all.
Adaptive: Layouts transform across 5 window size classes. Components resize, reposition, and change form factor responsively.
Expressive: Shape morphing, spring physics, and emphasized typography create moments of delight without sacrificing usability.

Current Updates: Google I/O 2026

Material's Google I/O 2026 update reinforces a Compose-first Android path and expands expressive/adaptive guidance:

Material Android is Compose-first: For new Android work, prefer Jetpack Compose Material3 for the latest components, expressive APIs, adaptive scaffolds, and Styles API integration. Android Views may remain necessary in existing apps, but they should not be treated as the default path for new Material 3 implementations.
Expressive layout system: Use an expressive layout scaffold to adapt screens across mobile, desktop, foldables, watches, XR, and other spatial form factors. Start from adaptive scaffolds/window size classes instead of fixed phone-first layouts.
8dp spacing system: Apply spacing tokens for margins, padding, and gaps so layouts and components can adapt programmatically to device type and density.
New/updated expressive components: Lists, menus, search, and search app bars have refreshed expressive guidance, with Jetpack Compose as the primary implementation target.
Watches and XR: Watches emphasize physics-based motion, arc text, and edge-hugging containers. XR emphasizes spatial panels and depth-based elevation.

Key differences from MD2:

Tonal surfaces replace elevation shadows as the primary depth cue
Dynamic color generates full schemes from a single seed color
Fully rounded corners by default (not slightly rounded)
Spring-based motion physics replace fixed easing curves for components
3 levels of user-controlled contrast (standard/medium/high)

Relationship with frontend-design skill: When both skills are active, MD3 provides the design system (tokens, components, layout rules) and frontend-design provides creative direction within those constraints. MD3 rules take precedence for component structure and token usage. Note: Roboto/Roboto Flex IS the correct default typeface in MD3 — the frontend-design guidance to avoid Roboto does not apply when implementing MD3.

Decision Tree

What are you building?

Full app scaffold        → See "Common Patterns: App Shell" + references/layout-and-responsive.md
Single component         → See "Component Quick Reference" table → references/component-catalog.md
Custom theme             → See references/theming-and-dynamic-color.md
Form / input layout      → See references/component-catalog.md § Input Components
Navigation structure     → See references/navigation-patterns.md
Data display             → See references/component-catalog.md § Data Display

What platform?

Jetpack Compose          → Primary: androidx.compose.material3, MaterialTheme, references/*
Flutter                  → useMaterial3: true in ThemeData, ColorScheme.fromSeed()
Web (vanilla JS)         → @material/web (limited; maintenance mode) + CSS custom properties
Web (React/Vue/Svelte)   → CSS custom properties + wrapper components (no official React lib)
Web (CSS-only)           → MD3 token values as CSS custom properties (no <md-*> elements)

Design Token System

All MD3 tokens use the md.sys namespace. Jetpack Compose maps roles to MaterialTheme.colorScheme, MaterialTheme.typography, and MaterialTheme.shapes (same semantic roles as the spec). On the web, these map to CSS custom properties (--md-sys-*):

Color Tokens (`--md-sys-color-*`)

Token	Purpose
`primary`	High-emphasis fills, text, icons against surface
`on-primary`	Text/icons on primary
`primary-container`	Standout fill for key components (FAB, etc.)
`on-primary-container`	Text/icons on primary-container
`secondary` / `on-secondary`	Less prominent accents
`secondary-container` / `on-secondary-container`	Recessive components (tonal buttons)
`tertiary` / `on-tertiary`	Contrasting accents
`tertiary-container` / `on-tertiary-container`	Complementary containers
`error` / `on-error`	Error states (static — doesn't change with dynamic color)
`error-container` / `on-error-container`	Error container fills
`surface`	Default background
`on-surface`	Text/icons on any surface
`on-surface-variant`	Lower-emphasis text/icons on surface
`surface-container-lowest`	Lowest-emphasis container
`surface-container-low`	Low-emphasis container
`surface-container`	Default container (nav areas)
`surface-container-high`	High-emphasis container
`surface-container-highest`	Highest-emphasis container
`surface-dim` / `surface-bright`	Maintain relative brightness across light/dark
`inverse-surface` / `inverse-on-surface` / `inverse-primary`	Contrasting elements (snackbars)
`outline`	Important boundaries (text field borders)
`outline-variant`	Decorative elements (dividers)

Full details: references/color-system.md

Typography Tokens (`--md-sys-typescale-*`)

Scale	Sizes	Use
Display	L / M / S	Hero text, large numbers
Headline	L / M / S	Section headers
Title	L / M / S	Smaller headers, card titles
Body	L / M / S	Paragraph text, descriptions
Label	L / M / S	Buttons, chips, captions

Each style has tokens for: -font, -weight, -size, -line-height, -tracking Plus 15 emphasized variants (higher weight) via --md-sys-typescale-emphasized-*

Full details: references/typography-and-shape.md

Shape Tokens (`--md-sys-shape-corner-*`)

Token	Value	Example components
`none`	0dp	—
`extra-small`	4dp	Chips, snackbars
`small`	8dp	Text fields, menus
`medium`	12dp	Cards
`large`	16dp	FABs, navigation drawer
`large-increased`	20dp	(Expressive)
`extra-large`	28dp	Dialogs, bottom sheets
`extra-large-increased`	32dp	(Expressive)
`extra-extra-large`	48dp	(Expressive)
`full`	9999px	Buttons, chips, badges

Elevation Levels

Level	DP	Tonal offset	Use
0	0dp	None	Flat surfaces, most components at rest
1	1dp	+5% primary	Elevated cards, modal sheets
2	3dp	+8% primary	Menus, nav bar, scrolled app bar
3	6dp	+11% primary	FAB, dialogs, search, date/time pickers
4	8dp	+12% primary	(hover/focus increase only)
5	12dp	+14% primary	(hover/focus increase only)

Elevation in MD3 is communicated through tonal surface color, not shadows. Shadows are only used when needed for additional protection against busy backgrounds.

Motion

MD3 Expressive (May 2025) introduced spring-based motion physics for components. The legacy easing/duration system is still used for transitions (enter/exit/shared-axis):

Easing	Duration	Transition type
Emphasized	500ms	Begin and end on screen
Emphasized decelerate	400ms	Enter the screen
Emphasized accelerate	200ms	Exit the screen
Standard	300ms	Begin and end on screen (utility)
Standard decelerate	250ms	Enter screen (utility)
Standard accelerate	200ms	Exit screen (utility)

CSS easing values:

Emphasized: cubic-bezier(0.2, 0, 0, 1)
Emphasized decelerate: cubic-bezier(0.05, 0.7, 0.1, 1)
Emphasized accelerate: cubic-bezier(0.3, 0, 0.8, 0.15)
Standard: cubic-bezier(0.2, 0, 0, 1)
Standard decelerate: cubic-bezier(0, 0, 0, 1)
Standard accelerate: cubic-bezier(0.3, 0, 1, 1)

Component Quick Reference

Component	Web Element	Key Variants	Category
Button	`md-filled-button`, `md-outlined-button`, `md-text-button`, `md-elevated-button`, `md-filled-tonal-button`	Filled, Outlined, Text, Elevated, Tonal; 5 sizes (XS–XL); toggle	Actions
Button group	`md-button-group`	Standard, connected	Actions
Extended FAB	`md-extended-fab`	Surface, Primary, Secondary, Tertiary	Actions
FAB	`md-fab`	Small, Medium, Large	Actions
FAB menu	—	—	Actions
Icon button	`md-icon-button`, `md-filled-icon-button`, `md-filled-tonal-icon-button`, `md-outlined-icon-button`	Standard, Filled, Filled Tonal, Outlined	Actions
Segmented button	—	Single-select, Multi-select	Actions
Split button	—	—	Actions
Badge	—	Small (dot), Large (count)	Communication
Loading indicator	—	Linear, Circular	Communication
Progress indicator	`md-linear-progress`, `md-circular-progress`	Linear, Circular; determinate/indeterminate	Communication
Snackbar	—	Single-line, Two-line, Action	Communication
Tooltip	—	Plain, Rich	Communication
Card	—	Filled, Outlined, Elevated	Containment
Carousel	—	Multi-browse, Uncontained, Hero	Containment
Dialog	`md-dialog`	Basic, Full-screen	Containment
Bottom sheet	—	Standard, Modal	Sheets
Side sheet	—	Standard, Modal	Sheets
Divider	`md-divider`	Full-width, Inset	Containment
Checkbox	`md-checkbox`	—	Input
Chips	`md-chip-set`, `md-assist-chip`, `md-filter-chip`, `md-input-chip`, `md-suggestion-chip`	Assist, Filter, Input, Suggestion	Input
Date picker	—	Docked, Modal, Range	Input
Menu	`md-menu`, `md-menu-item`	—	Input
Radio button	`md-radio`	—	Input
Slider	`md-slider`	Continuous, Discrete, Range	Input
Switch	`md-switch`	With/without icon	Input
Text field	`md-filled-text-field`, `md-outlined-text-field`	Filled, Outlined	Input
Time picker	—	Docked, Modal	Input
App bar (top)	—	Center-aligned, Small, Medium, Large	Navigation
Navigation bar	`md-navigation-bar`	—	Navigation
Navigation drawer	`md-navigation-drawer`	Standard, Modal	Navigation
Navigation rail	—	—	Navigation
Search	—	Search bar, Search view	Navigation
Tabs	`md-tabs`, `md-primary-tab`, `md-secondary-tab`	Primary, Secondary	Navigation
Toolbar	—	—	Navigation
List	`md-list`, `md-list-item`	One-line, Two-line, Three-line	Data Display

Note: Components marked with — for web element don't have @material/web implementations yet. Use CSS custom properties with standard HTML for these. Compose mappings and examples live in references/component-catalog.md.

Full component details with code examples: references/component-catalog.md

Jetpack Compose (primary)

Use androidx.compose.material3 with MaterialTheme and Material 3 composables (Scaffold, Button, NavigationBar, top app bars, etc.).

Theming: MaterialTheme(colorScheme = …, typography = …, shapes = …). Prefer dynamicLightColorScheme / dynamicDarkColorScheme on Android 12+ (API 31+) when dynamic color is desired; otherwise lightColorScheme / darkColorScheme or generated theme code from Material Theme Builder.
Adaptive UI: Window size classes, list-detail and supporting-pane layouts, foldables — see references/layout-and-responsive.md and references/navigation-patterns.md.
Edge-to-edge & insets: Lay out content with WindowInsets / scaffold padding so bars and IME behave correctly — see references/layout-and-responsive.md.
Experimental APIs: Some Material 3 APIs require @OptIn(ExperimentalMaterial3Api::class) or expressive opt-ins; match your BOM and compiler.

MaterialTheme(
    colorScheme = colorScheme, // from dynamicLightColorScheme / lightColorScheme / etc.
    typography = Typography(),
    shapes = Shapes(),
) {
    // M3 content — prefer references for Scaffold, navigation, text fields
}

Web (limited): @material/web

Important: Per Material Design 3 for Web, Material Web Components are in maintenance mode and M3 Expressive is not implemented on Web. Use @material/web for token-backed web UIs when appropriate, but do not treat it as equivalent to Compose for current Expressive features.

Setup

npm install @material/web

Import Components Individually

Always import only the components you use — importing the entire package bloats the bundle:

// Good — individual imports
import '@material/web/button/filled-button.js';
import '@material/web/button/outlined-button.js';
import '@material/web/textfield/outlined-text-field.js';
import '@material/web/icon/icon.js';

// Bad — never do this
import '@material/web'; // imports everything

Basic Usage

<!DOCTYPE html>
<html lang="en">
<head>
  <meta charset="UTF-8">
  <meta name="viewport" content="width=device-width, initial-scale=1.0">
  <link href="https://fonts.googleapis.com/css2?family=Roboto+Flex:wght@400;500;700&display=swap" rel="stylesheet">
  <link href="https://fonts.googleapis.com/icon?family=Material+Symbols+Outlined" rel="stylesheet">
</head>
<body>
  <md-filled-button>Get started</md-filled-button>
  <md-outlined-text-field label="Email" type="email"></md-outlined-text-field>

  <script type="module">
    import '@material/web/button/filled-button.js';
    import '@material/web/textfield/outlined-text-field.js';
  </script>
</body>
</html>

Theming with CSS Custom Properties

Apply a custom theme by setting CSS custom properties on :root or any ancestor:

:root {
  /* Color scheme (generate with @material/material-color-utilities) */
  --md-sys-color-primary: #6750A4;
  --md-sys-color-on-primary: #FFFFFF;
  --md-sys-color-primary-container: #EADDFF;
  --md-sys-color-on-primary-container: #21005D;
  --md-sys-color-secondary: #625B71;
  --md-sys-color-on-secondary: #FFFFFF;
  --md-sys-color-secondary-container: #E8DEF8;
  --md-sys-color-on-secondary-container: #1D192B;
  --md-sys-color-surface: #FEF7FF;
  --md-sys-color-on-surface: #1D1B20;
  --md-sys-color-surface-container: #F3EDF7;
  --md-sys-color-outline: #79747E;
  --md-sys-color-outline-variant: #CAC4D0;

  /* Typography */
  --md-sys-typescale-body-large-font: 'Roboto Flex', sans-serif;
  --md-sys-typescale-body-large-size: 1rem;
  --md-sys-typescale-body-large-weight: 400;
  --md-sys-typescale-body-large-line-height: 1.5rem;

  /* Shape */
  --md-sys-shape-corner-full: 9999px;
  --md-sys-shape-corner-medium: 12px;
}

Component-Level Overrides

Override individual component tokens for specific customization:

md-filled-button {
  --md-filled-button-container-color: var(--md-sys-color-primary);
  --md-filled-button-label-text-color: var(--md-sys-color-on-primary);
  --md-filled-button-container-shape: var(--md-sys-shape-corner-full);
  --md-filled-button-container-height: 40px;
}

md-outlined-text-field {
  --md-outlined-text-field-container-shape: var(--md-sys-shape-corner-small);
  --md-outlined-text-field-focus-outline-color: var(--md-sys-color-primary);
}

Dark Theme

Apply dark theme by overriding color tokens on a class or media query:

@media (prefers-color-scheme: dark) {
  :root {
    --md-sys-color-primary: #D0BCFF;
    --md-sys-color-on-primary: #381E72;
    --md-sys-color-primary-container: #4F378B;
    --md-sys-color-on-primary-container: #EADDFF;
    --md-sys-color-surface: #141218;
    --md-sys-color-on-surface: #E6E0E9;
    --md-sys-color-surface-container: #211F26;
    --md-sys-color-outline: #938F99;
    --md-sys-color-outline-variant: #49454F;
  }
}

Full theming guide: references/theming-and-dynamic-color.md

Common Patterns

App Shell

Standard MD3 app with responsive navigation + top app bar + content area:

<div class="md3-app">
  <nav class="md3-nav-rail" aria-label="Main navigation">
    <!-- Navigation rail for medium+ screens -->
    <md-fab size="small" aria-label="Compose">
      <md-icon slot="icon">edit</md-icon>
    </md-fab>
    <md-navigation-bar>
      <md-navigation-tab label="Home">
        <md-icon slot="active-icon">home</md-icon>
        <md-icon slot="inactive-icon">home</md-icon>
      </md-navigation-tab>
      <md-navigation-tab label="Search">
        <md-icon slot="active-icon">search</md-icon>
        <md-icon slot="inactive-icon">search</md-icon>
      </md-navigation-tab>
    </md-navigation-bar>
  </nav>
  <main class="md3-content">
    <header class="md3-top-app-bar">
      <h1 class="md3-top-app-bar__title" style="font: var(--md-sys-typescale-title-large)">
        Page Title
      </h1>
    </header>
    <div class="md3-body">
      <!-- Content here -->
    </div>
  </main>
</div>

.md3-app {
  display: flex;
  min-height: 100vh;
  background: var(--md-sys-color-surface);
  color: var(--md-sys-color-on-surface);
}

.md3-nav-rail {
  width: 80px;
  background: var(--md-sys-color-surface);
  border-right: 1px solid var(--md-sys-color-outline-variant);
  display: flex;
  flex-direction: column;
  align-items: center;
  padding-top: 12px;
  gap: 12px;
}

.md3-content {
  flex: 1;
  display: flex;
  flex-direction: column;
}

.md3-top-app-bar {
  height: 64px;
  padding: 0 16px;
  display: flex;
  align-items: center;
  background: var(--md-sys-color-surface);
}

.md3-body {
  padding: 24px;
  flex: 1;
}

/* Responsive: switch to bottom nav on compact */
@media (max-width: 599px) {
  .md3-app { flex-direction: column; }
  .md3-nav-rail {
    order: 1;
    width: 100%;
    flex-direction: row;
    justify-content: center;
    border-right: none;
    border-top: 1px solid var(--md-sys-color-outline-variant);
    padding: 0;
  }
}

Card Grid

<div class="md3-card-grid">
  <div class="md3-card md3-card--outlined">
    <img src="image.jpg" alt="Description" class="md3-card__media">
    <div class="md3-card__content">
      <h3 style="font: var(--md-sys-typescale-title-medium)">Card Title</h3>
      <p style="font: var(--md-sys-typescale-body-medium); color: var(--md-sys-color-on-surface-variant)">
        Supporting text for this card.
      </p>
    </div>
    <div class="md3-card__actions">
      <md-text-button>Learn more</md-text-button>
      <md-filled-tonal-button>Action</md-filled-tonal-button>
    </div>
  </div>
</div>

.md3-card-grid {
  display: grid;
  grid-template-columns: repeat(auto-fill, minmax(300px, 1fr));
  gap: 16px;
}

.md3-card--outlined {
  border: 1px solid var(--md-sys-color-outline-variant);
  border-radius: var(--md-sys-shape-corner-medium, 12px);
  background: var(--md-sys-color-surface);
  overflow: hidden;
}

.md3-card__content { padding: 16px; }
.md3-card__actions { padding: 8px 16px 16px; display: flex; gap: 8px; justify-content: flex-end; }
.md3-card__media { width: 100%; aspect-ratio: 16/9; object-fit: cover; }

Form Layout

<form class="md3-form">
  <md-outlined-text-field label="Full name" required></md-outlined-text-field>
  <md-outlined-text-field label="Email" type="email" required></md-outlined-text-field>
  <md-outlined-text-field label="Message" type="textarea" rows="4"></md-outlined-text-field>
  <div class="md3-form__actions">
    <md-text-button type="reset">Cancel</md-text-button>
    <md-filled-button type="submit">Submit</md-filled-button>
  </div>
</form>

.md3-form {
  display: flex;
  flex-direction: column;
  gap: 16px;
  max-width: 560px;
}

.md3-form__actions {
  display: flex;
  gap: 8px;
  justify-content: flex-end;
  margin-top: 8px;
}

More patterns: references/navigation-patterns.md, references/layout-and-responsive.md

Anti-Patterns

Never do these when implementing MD3:

Mix MD2 and MD3 libraries: Don't use @material/mdc-* (MD2) alongside @material/web (MD3). They have incompatible APIs and styling.
Hardcode colors: Always use var(--md-sys-color-*) tokens, never raw hex/rgb values. Hardcoded colors break dynamic theming, dark mode, and contrast adjustment.
Ignore tonal pairing: Only combine colors in their intended pairs (e.g., primary + on-primary, surface-container + on-surface). Arbitrary pairings break contrast in dynamic color and high contrast modes.
Use outline for dividers: Use outline-variant for dividers. outline is for important boundaries like text field borders.
Import all of @material/web: Always import individual component modules. Barrel imports include every component and destroy bundle size.
Use border-radius directly: Use shape tokens (var(--md-sys-shape-corner-medium)) so shapes stay consistent with theming.
Use shadows for elevation by default: MD3 communicates elevation through tonal surface color, not shadows. Only add shadows when elements need extra separation from busy backgrounds.
Apply frontend-design "avoid Roboto" rule: On Android, Roboto is the default Material typeface; web often uses Roboto or Roboto Flex with MD3 tokens. Replace only when intentionally customizing the type scale.
Assume SSR compatibility: @material/web uses Web Components (custom elements) which require JavaScript to render. They won't produce meaningful HTML in SSR without additional hydration strategies.
Ignore foldables and large screens: MD3 is designed for all screen sizes. Don't ship phone-only layouts — use canonical layouts, multi-pane at 600dp+, and test on foldable/tablet emulators. Place no interactive content across the fold/hinge.
Stretch content to fill wide screens: On Large (1200dp+) and Extra-large (1600dp+) windows, constrain content to a max width (840–1040dp). Endless-width text lines are unreadable.

Platform Notes

Flutter

MaterialApp(
  theme: ThemeData(
    useMaterial3: true,
    colorScheme: ColorScheme.fromSeed(seedColor: Colors.deepPurple),
  ),
);

Jetpack Compose

See Jetpack Compose (primary) above. Use LocalContext.current with dynamicLightColorScheme / dynamicDarkColorScheme only when Build.VERSION.SDK_INT >= Build.VERSION_CODES.S and dynamic color is enabled; otherwise supply static light/dark schemes.

Component Name Mapping

Concept	Web	Flutter	Compose
Filled button	`md-filled-button`	`FilledButton`	`Button`
Outlined text field	`md-outlined-text-field`	`OutlinedTextField`	`OutlinedTextField`
FAB	`md-fab`	`FloatingActionButton`	`FloatingActionButton`
Navigation bar	`md-navigation-bar`	`NavigationBar`	`NavigationBar`
Switch	`md-switch`	`Switch`	`Switch`

M3 Expressive (May 2025)

The Expressive update adds visual richness while maintaining usability. Availability differs by platform — do not assume one stack implements everything.

Capability	Jetpack Compose	Flutter	Web (`@material/web`)
Expressive layout scaffold / adaptive layout	Compose-first via Material3 adaptive APIs and window size classes	Use Flutter adaptive/layout primitives	CSS/container queries/manual layout; no Material Web parity
8dp spacing system	Use design tokens / `Dp` spacing constants; keep margins, padding, and gaps adaptive	Use theme spacing constants	CSS custom properties / design tokens
Expressive lists, menus, search, search app bar	Primary target per current Material guidance; check BOM and opt-ins	Check current Flutter Material docs	Spec-aligned custom implementation; `@material/web` is maintenance-only
Spring / motion physics	Supported in Material 3 (see `MotionScheme`, expressive APIs per BOM)	Varies by Flutter Material version	Not in Material Web; use easing/duration or custom motion
Emphasized typography	Via theme / type scale	Via theme	Token/CSS only; no full Expressive component set
Shape morphing	Compose-first in Google’s expressive rollout	Check current Flutter docs	Not in `@material/web`
New button sizes (XS–XL), toggle	Follow Compose Material3 components	Follow Flutter MD3	Height/CSS approximations only
Extra corner tokens (e.g. large-increased)	`MaterialTheme.shapes` / tokens	Theme shapes	CSS `--md-sys-shape-*`
3 contrast levels	Scheme builders / system	Plugins / manual	`SchemeContent` contrast parameter in JS utilities
Watches / XR form factors	Use Compose/Wear/XR-specific guidance where available	Platform-specific	Web/spatial UI custom implementation

Web: Material Web is maintenance-only; M3 Expressive is not on Web. Use CSS easing/duration tokens as fallback for motion, not spring parity.

Legacy easing/duration remains valid for transitions (enter/exit/shared-axis) where the spec still references them; see the Motion table below.

MD3 Compliance Audit

When invoked with audit as the argument (e.g., /material-3 audit), or when asked to audit/review MD3 compliance, analyze the target app or page and produce a compliance report.

Audit Procedure

Identify the target: The user provides a URL (use browser tools to inspect), file paths (read source), or a running app.
Inspect the following categories and score each 0–10:

Category	What to check
Color tokens	Web: `--md-sys-color-` / generated CSS. Compose:* `MaterialTheme.colorScheme` roles (no arbitrary `Color(...)` for surfaces without reason). Proper tonal pairing (`onX` on `X`). Dark theme. Flutter: `ColorScheme` roles.
Typography	MD3 type scale: Compose `MaterialTheme.typography`; web typescale tokens; correct roles (Display, Headline, Title, Body, Label).
Shape	Compose `MaterialTheme.shapes` / component `Shape`; web `var(--md-sys-shape-*)`. Buttons: full; cards: medium; avoid magic numbers.
Elevation	Tonal elevation (`Surface` tonal/shadow as appropriate). Web: hover/focus where relevant.
Components	Compose: Material3 composables (`Button`, `Scaffold`, etc.). Web: `@material/web` or spec-aligned HTML/CSS. Correct variants.
Layout	Canonical layouts; Compose window size class / adaptive APIs; readable max width on large widths; foldable hinge avoidance.
Navigation	Bar / rail / drawer / drawers+Compose `NavHost` patterns per size class; predictive back where applicable.
Motion	Compose `MotionScheme` / expressive APIs when used; transitions may still use easing/duration. Web: CSS motion tokens fallback.
Accessibility	MD3 roles help, but verify contrast: UI components often need 3:1 for large text/borders and 4.5:1 for normal text (WCAG 2.x). TalkBack/semantics (Compose), focus order, touch targets (~48dp). Web: ARIA, keyboard.
Theming	Compose: `MaterialTheme` + light/dark/dynamic as designed. Web: CSS custom properties on `:root` or subtree. Flutter: `ThemeData` + `ColorScheme`.

Generate the report:

# MD3 Compliance Audit Report

Target: [URL or file path]
Date: [date]
Overall Score: [X/100]

## Scores by Category
| Category       | Score | Status |
|----------------|-------|--------|
| Color tokens   | X/10  | [pass/warn/fail] |
| Typography     | X/10  | [pass/warn/fail] |
| Shape          | X/10  | [pass/warn/fail] |
| Elevation      | X/10  | [pass/warn/fail] |
| Components     | X/10  | [pass/warn/fail] |
| Layout         | X/10  | [pass/warn/fail] |
| Navigation     | X/10  | [pass/warn/fail] |
| Motion         | X/10  | [pass/warn/fail] |
| Accessibility  | X/10  | [pass/warn/fail] |
| Theming        | X/10  | [pass/warn/fail] |

## Critical Issues
[List items scoring 0-3 with specific file:line references and fixes]

## Warnings
[List items scoring 4-6 with recommendations]

## Passing
[List items scoring 7-10 with notes on what's done well]

## Recommended Fixes (Priority Order)
1. [Most impactful fix first]
2. ...

Audit Methods

For a live URL (browser or devtools):

Inspect computed styles and CSS variables (--md-sys-*)
Resize viewport or use responsive mode for breakpoints
Capture screenshots at key widths if helpful

For source code (file paths provided):

Compose/Kotlin: .kt files — MaterialTheme, composables, Color(0x…) abuse, hard-coded Dp, missing Modifier.semantics where needed
Flutter: .dart — ThemeData, ColorScheme
Web: HTML/JSX/Vue/Svelte; CSS/SCSS for tokens
Check web imports for @material/web vs @material/mdc-* (MD2)

Quick checks (adapt paths to your stack):

# Web: hardcoded colors
grep -rn '#[0-9a-fA-F]\{3,8\}' --include='*.css' --include='*.scss'

# Compose: raw Color(...) audits (sample — tune for your codebase)
grep -rn 'Color(0x' --include='*.kt'

# MD2 on web
grep -rn '@material/mdc-' --include='*.js' --include='*.ts'

Browser automation (if your environment exposes MCP browser tools): navigate, snapshot DOM/CSS variables, resize for breakpoints — optional, not required.

Scoring Guide

9-10: Fully MD3 compliant, uses correct tokens and patterns
7-8: Mostly compliant, minor issues (e.g., a few hardcoded values)
4-6: Partially compliant, some MD3 patterns but significant gaps
1-3: Major violations, mostly non-MD3 or MD2 patterns
0: Not applicable or completely absent

Status thresholds: pass (7+), warn (4-6), fail (0-3)

Reference Documents

references/color-system.md — Color roles, tonal palettes, dynamic color, Compose + CSS mapping
references/typography-and-shape.md — Type scale, shape corners, elevation, motion, Expressive notes
references/component-catalog.md — Components: Compose + @material/web where applicable
references/navigation-patterns.md — Navigation selection, Compose-first adaptive patterns
references/layout-and-responsive.md — Breakpoints, canonical layouts, insets, foldables
references/theming-and-dynamic-color.md — Theming: Compose first, then Flutter and web

When it activates

Automatically, when a task matches. Use when: "material design", "MD3", "material you", "Jetpack Compose", "MaterialTheme", "material component", "md3 button".

About this skill

Automatic

Frontend / UI pack

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v5 — 8/30/2026

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

sentenz / skills

skills/material-3/SKILL.md

View on GitHub ↗

License Apache-2.0Imported 8/30/2026Version v1Bundled 6 files

Illustrative Want to see what sentenz/material-3 would put in front of your model? Preview the rendered body and check its trigger phrases — no LLM call, no signup.

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

Material Design 3

Philosophy

Current Updates: Google I/O 2026

Decision Tree

Design Token System

Color Tokens (--md-sys-color-*)

Typography Tokens (--md-sys-typescale-*)

Shape Tokens (--md-sys-shape-corner-*)

Elevation Levels

Motion

Component Quick Reference

Jetpack Compose (primary)

Web (limited): @material/web

Setup

Import Components Individually

Basic Usage

Theming with CSS Custom Properties

Component-Level Overrides

Dark Theme

Common Patterns

App Shell

Card Grid

Form Layout

Anti-Patterns

Platform Notes

Flutter

Jetpack Compose

Component Name Mapping

M3 Expressive (May 2025)

MD3 Compliance Audit

Audit Procedure

Audit Methods

Scoring Guide

Reference Documents

When it activates

SkillSafety

Source

Related skills

More in the registry

Color Tokens (`--md-sys-color-*`)

Typography Tokens (`--md-sys-typescale-*`)

Shape Tokens (`--md-sys-shape-corner-*`)