Ready to supercharge your AI agents?

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

Cross Browser Testing · petrkindlmann | DecimalAI

Testingby petrkindlmann

Cross Browser Testing

Design analytics-driven browser test matrices and execute cross-browser tests. Covers BrowserStack/Sauce Labs configuration, Playwright browser channels, common cross-browser CSS/JS divergences, a known-issues documentation log, and progressive enhancement validation.

Maintained

SkillScore

SkillScore82/100

Test results 82

Safety

3/3 safety checks passed

Code scan
Intent
Content

What it changes

45%82%of test cases passed +36 pts

Without skill10 of 22

With skill18 of 22

8 fixed10 passed either way4 still failing0 broke

Measured on gemini-3.8-flash11/2/2026v1

Track record

Agent runs: 0; All time

See one task, with and without the skill

One test taskWe noticed that animated scrolling to in-page anchors via CSS `scroll-behavior: smooth` fails or behaves inconsistently on older WebKit engines. If we record this in our known browser issues log, what should the test assertion look like?

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 petrkindlmann-cross-browser-testing
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("petrkindlmann-cross-browser-testing")
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 petrkindlmann-cross-browser-testing

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/petrkindlmann-cross-browser-testing/SKILL.md https://app.decimal.ai/s/petrkindlmann-cross-browser-testing/SKILL.md

Copy SKILL.md

Save it in your project, as

.claude/skills/petrkindlmann-cross-browser-testing/SKILL.md

This gets SKILL.md only. The skill's 3 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 petrkindlmann-cross-browser-testing

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("petrkindlmann-cross-browser-testing")
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/petrkindlmann-cross-browser-testing/SKILL.md

Give it to your agent

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

skill = Path(".agents/skills/petrkindlmann-cross-browser-testing/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 3 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 petrkindlmann-cross-browser-testing

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("petrkindlmann-cross-browser-testing")
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/petrkindlmann-cross-browser-testing/SKILL.md

Give it to your agent

from pathlib import Path
from agents import Agent, Runner

skill = Path(".agents/skills/petrkindlmann-cross-browser-testing/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 3 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 petrkindlmann-cross-browser-testing

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("petrkindlmann-cross-browser-testing")
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/petrkindlmann-cross-browser-testing/SKILL.md

Give it to your agent

from pathlib import Path
from openai import OpenAI

client = OpenAI()
skill = Path(".agents/skills/petrkindlmann-cross-browser-testing/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 3 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 petrkindlmann-cross-browser-testing

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("petrkindlmann-cross-browser-testing")
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/petrkindlmann-cross-browser-testing/SKILL.md

Give it to your agent

from pathlib import Path
import anthropic

client = anthropic.Anthropic()
skill = Path(".agents/skills/petrkindlmann-cross-browser-testing/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 3 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 petrkindlmann-cross-browser-testing

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("petrkindlmann-cross-browser-testing")
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/petrkindlmann-cross-browser-testing/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/petrkindlmann-cross-browser-testing/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 3 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 petrkindlmann-cross-browser-testing

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("petrkindlmann-cross-browser-testing")
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/petrkindlmann-cross-browser-testing/SKILL.md

Give it to your agent

from pathlib import Path
from langchain.agents import create_agent

skill = Path(".agents/skills/petrkindlmann-cross-browser-testing/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 3 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 petrkindlmann-cross-browser-testing

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("petrkindlmann-cross-browser-testing")
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/petrkindlmann-cross-browser-testing/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 3 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 petrkindlmann-cross-browser-testing

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("petrkindlmann-cross-browser-testing")
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/petrkindlmann-cross-browser-testing/SKILL.md

Give it to your agent

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

skill = Path(".agents/skills/petrkindlmann-cross-browser-testing/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 3 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 petrkindlmann-cross-browser-testing

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("petrkindlmann-cross-browser-testing")
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/petrkindlmann-cross-browser-testing/SKILL.md

Give it to your agent

from pathlib import Path
from pydantic_ai import Agent

skill = Path(".agents/skills/petrkindlmann-cross-browser-testing/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 3 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 petrkindlmann-cross-browser-testing

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("petrkindlmann-cross-browser-testing")
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/petrkindlmann-cross-browser-testing/SKILL.md

Give it to your agent

from pathlib import Path
from deepagents import create_deep_agent

skill = Path(".agents/skills/petrkindlmann-cross-browser-testing/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 3 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 petrkindlmann-cross-browser-testing

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("petrkindlmann-cross-browser-testing")
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/petrkindlmann-cross-browser-testing/SKILL.md

Give it to your agent

import asyncio
from pathlib import Path
from agent_framework import Agent

skill = Path(".agents/skills/petrkindlmann-cross-browser-testing/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 3 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 petrkindlmann-cross-browser-testing

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("petrkindlmann-cross-browser-testing")
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/petrkindlmann-cross-browser-testing/SKILL.md

Give it to your agent

from pathlib import Path
from strands import Agent

skill = Path(".agents/skills/petrkindlmann-cross-browser-testing/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 3 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 petrkindlmann-cross-browser-testing

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/petrkindlmann-cross-browser-testing/SKILL.md https://app.decimal.ai/s/petrkindlmann-cross-browser-testing/SKILL.md

Copy SKILL.md

Save it in your project, as

.agents/skills/petrkindlmann-cross-browser-testing/SKILL.md

This gets SKILL.md only. The skill's 3 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/petrkindlmann-cross-browser-testing/SKILL.md

Read the summary as JSON

curl -fsSL https://app.decimal.ai/s/petrkindlmann-cross-browser-testing.json

This gets SKILL.md only. The skill's 3 other files are not included.

Other ways to add it

Other setups docs

You don't need a DecimalAI account.

Version: v1, tested 11/2/2026
Source: petrkindlmann/qa-skills ↗
License: MIT

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

Quick Route

Situation	Go to
Need to decide which browsers to test	Browser Matrix Design
Already on Playwright, just add browsers	Playwright Browser Configuration → `references/playwright-and-cloud-config.md`
Need real Safari/Windows/older OS, not engines	Cloud Platform Setup
One browser misbehaves; want a test for it	Common Cross-Browser Issues + `browserName` branch in `references/testing-patterns.md`
Need to record a divergence so it is not re-debugged	Known-Issues Log
Pixel diffs / baseline thresholds	use `visual-testing`

Discovery Questions

Check .agents/qa-project-context.md first — if it exists, use it and skip anything already answered there. Then:

Target browsers from analytics: What do actual users use? Pull browser/OS data from your analytics tool. Testing browsers nobody uses is waste; missing a browser 15% of users rely on is a bug.
Desktop and mobile? Mobile Safari on iOS and Chrome on Android render differently than their desktop counterparts. Treat them as separate matrix entries.
Cloud platform: BrowserStack, Sauce Labs, LambdaTest, or local engines only? Cloud platforms provide real branded browsers and OSes; Playwright's bundled engines cover Chromium, Firefox, and WebKit (not Chrome/Safari themselves).
Progressive enhancement or pixel-perfect? Progressive enhancement accepts graceful degradation. Pixel-perfect demands identical rendering. The answer determines pass/fail criteria.
Existing Playwright config? If the project already uses Playwright, cross-browser testing is a configuration change, not a new tool.

Core Principles

Analytics-driven matrix. Test what your users actually use. A browser at 0.3% traffic does not need the same investment as one at 40%. Check analytics quarterly — browser share shifts.
Progressive enhancement over pixel-perfect. Identical rendering across all browsers is neither achievable nor necessary. Define what "works" means: core functionality operates, content is accessible, layout is usable. Visual differences in shadows, gradients, or animation timing are acceptable.
Safari and Firefox surface the most cross-browser bugs. Chrome-only testing catches Chrome bugs. Safari's WebKit engine and Firefox's Gecko engine have the most behavioral differences from Chromium. Prioritize them.
Test functionality, not rendering-engine internals. A cross-browser test should verify that the user can complete a task, not that a CSS property renders identically. Visual comparison tools handle pixel-level differences.
Engines are not brands. Playwright's WebKit is not Safari and its Chromium is not Chrome — they share an engine, not the shipped product (codecs, fonts, enterprise policy, update cadence all differ). Report "WebKit coverage," not "Safari coverage," unless you ran real Safari on a cloud grid.
One test, multiple browsers. Write tests once. Run them across browser configurations. Never duplicate test logic for different browsers.

Browser Matrix Design

Analytics-Based Methodology

Step 1: Export browser/OS data from analytics (last 90 days)
Step 2: Rank by session share
Step 3: Group into tiers
Step 4: Assign test coverage per tier
Step 5: Review quarterly

Tier System

Tier	Criteria	Coverage	When to run
P0	>10% traffic share	Full test suite	Every PR, every deploy
P1	3-10% traffic share	Smoke + critical paths	Nightly, pre-release
P2	1-3% traffic share	Smoke tests only	Weekly, pre-release
Skip	<1% traffic share	Not tested	Manual spot-check if reported

Example Matrix (derived from analytics)

## Browser Matrix — Q1 2026 (next-review: 2026-04-01)

| Browser | Version | Platform | Traffic % | Tier | Notes |
|---------|---------|----------|-----------|------|-------|
| Chrome | Latest | Windows | 34% | P0 | |
| Chrome | Latest | macOS | 12% | P0 | |
| Safari | Latest | macOS | 11% | P0 | WebKit-specific issues |
| Chrome | Latest | Android | 15% | P0 | Mobile viewport |
| Safari | Latest | iOS | 14% | P0 | Mobile Safari quirks |
| Firefox | Latest | Windows | 5% | P1 | Gecko rendering |
| Edge | Latest | Windows | 4% | P1 | Chromium-based but different UA/policy |
| Samsung Internet | Latest | Android | 3% | P1 | Chromium fork, lagging engine |
| Firefox | Latest | macOS | 1.5% | P2 | |
| Chrome | N-1 | Windows | 1.2% | P2 | Previous major version |

Version Coverage Strategy

Latest: Always test current stable release.
Latest - 1: Test previous major version only for P0 browsers where analytics show >1% on older versions.
Extended Support Release (ESR): Test Firefox ESR only if enterprise users are a significant segment.
Do not test: Beta/Canary/Nightly releases unless you are a browser vendor or building browser-facing tools.

Playwright Browser Configuration

Playwright ships three browser engines — Chromium, Firefox, WebKit — so no cloud platform is needed for basic engine-level coverage. This is engine coverage, not brand coverage: bundled WebKit ≠ Safari and bundled Chromium ≠ Chrome (see Core Principle 5). Define one project per matrix entry, map mobile devices via devices[...], and drive locally installed branded browsers with the channel option.

See references/playwright-and-cloud-config.md for the full playwright.config.ts project list, branded-channel snippets, and --project run commands.

When to use channels: When you need real branded behavior that differs from the bundled engine — installed Chrome (channel: 'chrome') or Edge (channel: 'msedge') for extension support, enterprise policy, or codecs. WebKit and Firefox have no channel option; they are always Playwright's bundled engines. Note the edge project in the config and the msedge channel snippet are illustrative alternatives, not two projects to merge — a config needs one edge project, not both.

page.screencast() (Playwright 1.59+, current in 1.60) captures annotated video of a cross-browser run — useful when a matrix failure needs human review across engines. For agent-driven re-runs and stepping through a failure, use --ui (UI mode) or --debug (Inspector); PWDEBUG=1 and --headed are the other real entry points. There is no --debug=cli flag.

Cloud Platform Setup

Cloud platforms (BrowserStack, Sauce Labs) provide real branded-browser/OS instances Playwright connects to over a CDP/Playwright WebSocket endpoint. Pass credentials and capabilities via environment variables, and keep the platform's playwrightVersion aligned with the Playwright version in package.json (currently 1.60.x — a client/server mismatch causes socket errors).

BrowserStack now recommends the npx browserstack-node-sdk runner plus a client.playwrightVersion capability (in addition to browserstack.playwrightVersion) to keep the client and grid sockets in lock-step. The raw wsEndpoint/CDP config below still works for direct connections; use the SDK path for new setups.

See references/playwright-and-cloud-config.md for the BrowserStack config (with the client.playwrightVersion cap), the Sauce Labs config, and the GitHub Actions parallel matrix that fans out across cloud browsers.

Common Cross-Browser Issues

Real divergences that surface in cross-browser testing, with detection patterns and fixes. The CSS workarounds and Playwright tests for each are in references/common-browser-issues.md, covering: partitioned cookies / CHIPS in iframes, <input type="date">, the Clipboard API, scroll-behavior, backdrop-filter, the <dialog> element, View Transitions, and Web Animations timing.

Modern Cross-Browser Gotchas (2026)

The classic Safari-laggard list is mostly resolved (flexbox gap, :has() shipping, same-document View Transitions are all Baseline). Today's real divergences:

Partitioned cookies / partitioned storage: Chrome's CHIPS (Partitioned attribute), Safari's ITP, and Firefox's State Partitioning each behave differently for embedded third-party contexts. Test third-party cookies in an iframe per engine, not just "the browser supports cookies." See the runnable per-engine iframe test in references/common-browser-issues.md.
:has() selector performance: Universally supported since 2023, but a :has()-heavy page can have very different style-recalc cost across engines. Watch list — profile if a page feels janky in one engine; visual-regression it in visual-testing.
View Transitions API: Same-document transitions are Baseline (Chrome 111, Safari 18, Firefox 144 — Oct 2025), so they are no longer a divergence. Cross-document transitions are still the gap: Chrome 126+, Safari 18.2+, Firefox behind a flag. Treat cross-document as progressive enhancement and verify the no-transition fallback.
WebDriver BiDi: Production-ready in Selenium 4, partially supported in Playwright. For new cross-runner projects, BiDi is the convergence point. Watch list.

Known-Issues Log

When a divergence is real and you cannot fix the app immediately, record it in a committed file (docs/browser-issues.md) so it is not re-debugged from scratch. The table is the artifact Done When checks for, and every row's test must assert the user outcome, not the CSS property:

| Affected browser | Repro | Workaround / fallback / ticket | Test asserts (user outcome, not CSS) |
|------------------|-------|--------------------------------|--------------------------------------|
| Safari (WebKit) ≤17 | scroll-behavior: smooth is partial | rely on anchor nav; no JS scroll dependency | anchor link puts heading in viewport (`toBeInViewport`) |
| Firefox ≤102 | backdrop-filter unsupported | -webkit- prefix + rgba background fallback | overlay readable; modal content visible |
| Firefox (current) | cross-document View Transitions flagged off | progressive enhancement; instant nav fallback | navigation completes; target page heading visible |

Keep one row per divergence. A row with no ticket and no fallback is an open bug, not a documented issue.

Testing Patterns

The core patterns and the rules that govern them:

Same test, multiple browsers — the default. Write the test once; configure projects to run it everywhere. Never duplicate test logic per browser.
Browser-specific test logic — branch on browserName only when behavior genuinely differs (the WebKit date-input fallback and Chromium-only clipboard permission are real cases). Rule: keep this rare. Many browser branches signal application compatibility bugs to fix, not work around.
Visual cross-browser comparison — toHaveScreenshot with a maxDiffPixelRatio tolerance; each browser project generates its own baseline (homepage-chromium.png, homepage-webkit.png, …). For threshold strategy and baseline management, use visual-testing.
Progressive enhancement validation — abort script requests (Chromium only) and verify core functionality still works via native HTML.

See references/testing-patterns.md for the runnable code for all four patterns.

Anti-Patterns

Testing only on Chrome. Chrome is the largest desktop share but uses the same engine as Edge, Opera, and Brave. Safari (WebKit) and Firefox (Gecko) surface the real cross-browser issues. Chrome-only testing gives false confidence.

Reporting WebKit/Chromium as Safari/Chrome. Bundled engines share rendering, not the shipped browser. Claiming "Safari coverage" off a WebKit project hides codec, font, and policy bugs that only real Safari shows.

Testing every browser equally. A browser at 1% traffic share does not need the same investment as one at 30%. Use the tier system to allocate effort proportionally.

Duplicating tests per browser. Write tests once, run them across browser projects via configuration. A checkout.chrome.spec.ts and checkout.safari.spec.ts with identical logic is the wrong shape.

browserName checks everywhere. Excessive browser branching in tests signals application compatibility issues. Fix the app, do not work around it in tests.

Pixel-perfect assertions without tolerance. Font rendering, anti-aliasing, and sub-pixel rounding differ between browsers and platforms. Use maxDiffPixelRatio or maxDiffPixels in visual comparisons.

Ignoring mobile browsers. Mobile Chrome and mobile Safari are not their desktop counterparts — different viewport behavior, touch handling, and CSS support. Test them as separate matrix entries.

Static browser matrix. Browser usage changes. A matrix based on data from two years ago is wrong. Review analytics quarterly and update the next-review date.

Documenting a divergence with no fallback or ticket. A known-issues row that lists no workaround and no open ticket is an undocumented bug pretending to be documented.

Failure Modes

Symptom	Likely cause	Fix or check
Cloud tests fail with a socket/handshake error	Grid Playwright version ≠ local	Set `playwrightVersion`/`client.playwrightVersion` to match `npx playwright --version`; use the `browserstack-node-sdk` runner
Clipboard test passes in Chromium, fails in Firefox/WebKit	`grantPermissions` only works in Chromium	Assert UI feedback (`Copied!`), not the clipboard API; gate `grantPermissions` on `browserName === 'chromium'`
Progressive-enhancement test errors in Firefox/WebKit	Script-abort route interception is Chromium-only	Gate the route on `browserName === 'chromium'`; skip the JS-disabled assertion elsewhere
WebKit project "passes" but real users on Safari report breakage	WebKit engine ≠ shipped Safari	Add a real-Safari row on a cloud grid for the affected flow
Visual baseline diff explodes for one browser only	Single baseline shared across browsers	Generate per-project baselines; each browser keeps its own `*-<project>.png`
`:has()`-heavy page janky in one engine only	Style-recalc cost differs by engine	Profile in that engine; reduce `:has()` scope; visual-regress in `visual-testing`

Done When

Browser matrix defined using real analytics data (last 90 days), with tier assignments (P0/P1/P2) documented and justified by traffic share, committed to a file carrying a dated next-review field.
Playwright project config (or BrowserStack/Sauce Labs config) reflects the matrix and runs P0 browsers on every PR; cloud configs pin playwrightVersion to match package.json.
docs/browser-issues.md exists with one row per known divergence: affected browser, repro, workaround/fallback or linked ticket, and the test that asserts the user outcome (not the CSS).
Common-divergence checklist (partitioned cookies, date inputs, clipboard, scroll behavior, backdrop-filter, <dialog>, View Transitions) has a test or a known-issues row for each item relevant to P0/P1 browsers.
A tracked issue exists for the next quarterly matrix review (or the matrix file's next-review date is in the future), so the refresh is not lost.

Related Skills

visual-testing — Owns pixel-level baseline strategy, threshold tables, and toHaveScreenshot config. Go there for how tolerant a screenshot diff should be; this skill only decides which browsers get a baseline.
playwright-automation — Core Playwright patterns, fixtures, and CI configuration that cross-browser testing builds on.
ci-cd-integration — Pipeline configuration for parallel browser-matrix execution and artifact collection.
mobile-testing — Device-farm and native/hybrid app testing (Appium/Detox); go there when the target is an app, not a browser viewport.
accessibility-testing — Cross-browser accessibility differences (screen-reader behavior, ARIA support) that overlap with this matrix.

Reference Files (in `references/`)

playwright-and-cloud-config.md — playwright.config.ts project list, branded channels, --project run commands, and BrowserStack (SDK + client.playwrightVersion)/Sauce Labs/CI matrix configs.
common-browser-issues.md — Per-engine partitioned-cookie iframe test, date-input WebKit fallback, clipboard, scroll behavior, backdrop-filter, <dialog>, View Transitions, and Web Animations.
testing-patterns.md — Same-test-multiple-browsers, browserName branching, visual comparison, and progressive-enhancement code.

When it activates

Automatically, when a task matches. Use when: "cross-browser," "browser matrix," "BrowserStack," "Safari issues," "browser compatibility," "Edge," "works in Chrome but not Safari." Not for: pixel-level baseline strategy and threshold tuning — use visual-testing; devi

About this skill

Automatic

QA & Test Engineering pack

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v8 — 10/26/2026

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

petrkindlmann / qa-skills

skills/cross-browser-testing/SKILL.md

View on GitHub ↗

License MITImported 10/26/2026Version v1Bundled 3 files

Illustrative Want to see what petrkindlmann/cross-browser-testing 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

Quick Route

Discovery Questions

Core Principles

Browser Matrix Design

Analytics-Based Methodology

Tier System

Example Matrix (derived from analytics)

Version Coverage Strategy

Playwright Browser Configuration

Cloud Platform Setup

Common Cross-Browser Issues

Modern Cross-Browser Gotchas (2026)

Known-Issues Log

Testing Patterns

Anti-Patterns

Failure Modes

Done When

Related Skills

Reference Files (in references/)

When it activates

SkillSafety

Source

Related skills

More in the registry

Reference Files (in `references/`)