Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →When the user wants to create AI-generated ad creative, test performance creative, manage creative fatigue, or optimize paid media with AI tools. Also use when the user mentions 'ad creative,' 'performance creative,' 'creative testing,' 'creative fatigue,' 'Meta ads,' 'Google ads,' 'TikTok ads,' 'AI ads,' 'ad budget,' 'ROAS,' 'Advantage+,' or 'Performance Max.' This skill covers AI-powered paid creative from generation through performance optimization.
SkillScore80/100
Test results 80
3/3 safety checks passed
16%80%of test cases passed +64 pts
Passed 20 of 25 test cases with the skill, and 4 of 25 without it.
17 fixed3 passed either way4 still failing1 broke
One test taskWhat is the maximum number of creative assets allowed per ad in TikTok Smart+, and what integrated tool enables AI automated translation, dubbing, and video concept generation?
Adds every file the skill has, pinned so you can update it. No account needed.
pip install "decimalai>=0.15"
decimalai skills add tech-leads-club-paid-creative-ai
claude --plugin-dir "$(git rev-parse --show-toplevel 2>/dev/null || pwd)/.agents"The same install from Python, for a notebook or a setup script.
pip install "decimalai>=0.15"from decimalai import skillpkg
# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("tech-leads-club-paid-creative-ai")
if result.exit_code:
raise SystemExit(result.problems)
print("Added. Start Claude Code with the command below.")claude --plugin-dir "$(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.
/plugin marketplace add decimal-labs/decimalai-claude-plugins
/plugin install decimalai@decimalai/decimalai:install tech-leads-club-paid-creative-aiThe plugin needs python3 and curl on your machine.
No Python or account. Claude Code picks it up in its next session.
curl -fsSL --create-dirs -o .claude/skills/tech-leads-club-paid-creative-ai/SKILL.md https://app.decimal.ai/s/tech-leads-club-paid-creative-ai/SKILL.md.claude/skills/tech-leads-club-paid-creative-ai/SKILL.mdThis gets SKILL.md only. The skill's 1 other file is not included.
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.
pip install "decimalai[claude-agent-sdk]"
decimalai skills add tech-leads-club-paid-creative-aiimport 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.
pip install "decimalai[claude-agent-sdk]"from decimalai import skillpkg
# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("tech-leads-club-paid-creative-ai")
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.
.agents/skills/tech-leads-club-paid-creative-ai/SKILL.mdimport asyncio
from pathlib import Path
from claude_agent_sdk import ClaudeAgentOptions, query
skill = Path(".agents/skills/tech-leads-club-paid-creative-ai/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 1 other file is not included.
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.
pip install "decimalai[openai-agents]"
decimalai skills add tech-leads-club-paid-creative-aifrom 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.
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.
pip install "decimalai[openai-agents]"from decimalai import skillpkg
# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("tech-leads-club-paid-creative-ai")
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.
.agents/skills/tech-leads-club-paid-creative-ai/SKILL.mdfrom pathlib import Path
from agents import Agent, Runner
skill = Path(".agents/skills/tech-leads-club-paid-creative-ai/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 1 other file is not included.
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.
pip install "decimalai[openai]"
decimalai skills add tech-leads-club-paid-creative-aiimport 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.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.
pip install "decimalai[openai]"from decimalai import skillpkg
# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("tech-leads-club-paid-creative-ai")
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.
.agents/skills/tech-leads-club-paid-creative-ai/SKILL.mdfrom pathlib import Path
from openai import OpenAI
client = OpenAI()
skill = Path(".agents/skills/tech-leads-club-paid-creative-ai/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].messageThis gets SKILL.md only. The skill's 1 other file is not included.
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.
pip install "decimalai[anthropic]"
decimalai skills add tech-leads-club-paid-creative-aiimport 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()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.
pip install "decimalai[anthropic]"from decimalai import skillpkg
# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("tech-leads-club-paid-creative-ai")
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.
.agents/skills/tech-leads-club-paid-creative-ai/SKILL.mdfrom pathlib import Path
import anthropic
client = anthropic.Anthropic()
skill = Path(".agents/skills/tech-leads-club-paid-creative-ai/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 1 other file is not included.
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.
pip install "decimalai[google-genai]"
decimalai skills add tech-leads-club-paid-creative-aifrom 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)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.
pip install "decimalai[google-genai]"from decimalai import skillpkg
# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("tech-leads-club-paid-creative-ai")
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.
.agents/skills/tech-leads-club-paid-creative-ai/SKILL.mdfrom pathlib import Path
from google import genai
from google.genai import types
client = genai.Client()
skill = Path(".agents/skills/tech-leads-club-paid-creative-ai/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 1 other file is not included.
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.
pip install "decimalai[langchain]" langchain-openai
decimalai skills add tech-leads-club-paid-creative-aifrom 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!"}]})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.
pip install "decimalai[langchain]" langchain-openaifrom decimalai import skillpkg
# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("tech-leads-club-paid-creative-ai")
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.
.agents/skills/tech-leads-club-paid-creative-ai/SKILL.mdfrom pathlib import Path
from langchain.agents import create_agent
skill = Path(".agents/skills/tech-leads-club-paid-creative-ai/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 1 other file is not included.
CrewAI's own skill loader reads the installed folder. No DecimalAI account needed.
pip install "decimalai>=0.16" "crewai>=1.15.27"
decimalai skills add tech-leads-club-paid-creative-aiimport 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.
pip install "decimalai>=0.16" "crewai>=1.15.27"from decimalai import skillpkg
# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("tech-leads-club-paid-creative-ai")
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.
.agents/skills/tech-leads-club-paid-creative-ai/SKILL.mdfrom 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 1 other file is not included.
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.
pip install "decimalai[adk]"
decimalai skills add tech-leads-club-paid-creative-aifrom 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.
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.
pip install "decimalai[adk]"from decimalai import skillpkg
# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("tech-leads-club-paid-creative-ai")
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.
.agents/skills/tech-leads-club-paid-creative-ai/SKILL.mdfrom pathlib import Path
from google.adk.agents import LlmAgent
skill = Path(".agents/skills/tech-leads-club-paid-creative-ai/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 1 other file is not included.
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.
pip install "decimalai[pydantic-ai]"
decimalai skills add tech-leads-club-paid-creative-aifrom 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.
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.
pip install "decimalai[pydantic-ai]"from decimalai import skillpkg
# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("tech-leads-club-paid-creative-ai")
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.
.agents/skills/tech-leads-club-paid-creative-ai/SKILL.mdfrom pathlib import Path
from pydantic_ai import Agent
skill = Path(".agents/skills/tech-leads-club-paid-creative-ai/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 1 other file is not included.
Deep Agents' own skill loader reads the installed folder. No DecimalAI account needed.
pip install "decimalai>=0.16" deepagents
decimalai skills add tech-leads-club-paid-creative-aiimport 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.
pip install "decimalai>=0.16" deepagentsfrom decimalai import skillpkg
# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("tech-leads-club-paid-creative-ai")
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.
.agents/skills/tech-leads-club-paid-creative-ai/SKILL.mdfrom pathlib import Path
from deepagents import create_deep_agent
skill = Path(".agents/skills/tech-leads-club-paid-creative-ai/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 1 other file is not included.
Its own skill loader reads the installed folder. No DecimalAI account needed.
pip install "decimalai>=0.16" agent-framework
decimalai skills add tech-leads-club-paid-creative-aiimport 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.
pip install "decimalai>=0.16" agent-frameworkfrom decimalai import skillpkg
# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("tech-leads-club-paid-creative-ai")
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.
.agents/skills/tech-leads-club-paid-creative-ai/SKILL.mdimport asyncio
from pathlib import Path
from agent_framework import Agent
skill = Path(".agents/skills/tech-leads-club-paid-creative-ai/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 1 other file is not included.
Strands' own skill loader reads the installed folder. No DecimalAI account needed.
pip install "decimalai>=0.16" strands-agents
decimalai skills add tech-leads-club-paid-creative-aifrom 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.
pip install "decimalai>=0.16" strands-agentsfrom decimalai import skillpkg
# Writes .agents/skills and .decimal/skills.toml. Safe to run twice.
result = skillpkg.add("tech-leads-club-paid-creative-ai")
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.
.agents/skills/tech-leads-club-paid-creative-ai/SKILL.mdfrom pathlib import Path
from strands import Agent
skill = Path(".agents/skills/tech-leads-club-paid-creative-ai/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 1 other file is not included.
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 tech-leads-club-paid-creative-aiNo Python or account. Cursor, Copilot, Cline and Warp read this folder by themselves.
curl -fsSL --create-dirs -o .agents/skills/tech-leads-club-paid-creative-ai/SKILL.md https://app.decimal.ai/s/tech-leads-club-paid-creative-ai/SKILL.md.agents/skills/tech-leads-club-paid-creative-ai/SKILL.mdThis gets SKILL.md only. The skill's 1 other file is not included.
Give its text to your model as instructions. No account or package needed.
This gets SKILL.md only. The skill's 1 other file is not included.
Raw files: /s/tech-leads-club-paid-creative-ai/SKILL.md · /s/tech-leads-club-paid-creative-ai.json
Fetch the file and give its text to your model as instructions. No account or package needed.
curl -fsSL https://app.decimal.ai/s/tech-leads-club-paid-creative-ai/SKILL.mdcurl -fsSL https://app.decimal.ai/s/tech-leads-club-paid-creative-ai.jsonThis gets SKILL.md only. The skill's 1 other file is not included.
You don't need a DecimalAI account.
Technical implementation, co
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.
You are a performance creative strategist who builds AI-powered ad creative systems across Meta, Google, TikTok, YouTube, and LinkedIn. You combine platform-native AI tools (Advantage+, Performance Max, Smart+) with generative AI production (Runway, Midjourney, Pika) to create, test, and scale ad creative that drives measurable ROAS.
Ask the user:
| Platform | AI System | What It Does | Best For |
|---|---|---|---|
| Meta | Advantage+ Creative | Auto-generates background variants, text overlays, aspect ratios from one asset | Scaling static + video across placements |
| Meta | GEM (Generative Ads Model) | Personalizes ad delivery per user, auto-enhances Reels creative | Reels-first campaigns, conversion lift |
| Performance Max (Gemini) | Generates headlines, descriptions, images from landing page + existing assets | Full-funnel cross-channel (Search, Display, YouTube, Shopping) | |
| TikTok | Smart+ | Module-level automation (targeting, budget, creative, placement) | TikTok-native performance campaigns |
| TikTok | Symphony Creative | Gen AI video creation, auto-enhancements, translation, dubbing | Multi-language, high-volume TikTok creative |
Advantage+ takes a single product image and generates multiple variants with different backgrounds, colors, and layouts optimized for every placement (Feed, Stories, Reels, Audience Network).
Setup best practices:
What Advantage+ automates:
What you still control:
Performance Max uses Gemini to generate headlines, descriptions, images, and video assets from your landing page and existing creative. Campaigns run across Search, Display, YouTube, Discover, Gmail, and Maps simultaneously.
Asset group requirements (maximize all):
PMax optimization rules:
Smart+ unifies manual and automated experiences into a single flow with module-by-module control over targeting, budget, creative, and placement.
Smart+ creative features:
Symphony creative tools:
| Tool | Type | Best For | Price Range | Key Strength |
|---|---|---|---|---|
| Midjourney | Image | Hero images, product shots, lifestyle scenes | $10-$120/mo | Photorealistic quality, text rendering (v6.1+) |
| DALL-E 3 | Image | Rapid ideation, text-heavy creatives | Via ChatGPT Plus $20/mo | Strong text rendering, fast iteration |
| Adobe Firefly | Image | Brand-safe commercial use, editing workflows | $10/mo (unlimited) | Commercially licensed, Photoshop integration |
| Runway Gen-4 | Video | Product demos, B-roll, character-consistent ads | $12-$76/mo | Character consistency across shots |
| Pika | Video | Quick social video, animated product shots | $8-$58/mo | Cost-efficient, lively output |
| Sora | Video | High-quality brand video, cinematic ads | Via ChatGPT Plus | Cinematic quality, longer clips |
| Creatify | Full pipeline | URL-to-video ads, AI avatars, batch creation | $33-$69/mo | 1,000+ avatars, 29 languages, batch rendering |
| Arcads | UGC-style | Conversion-focused talking-head ads | $110-$220/mo | Most authentic AI UGC, natural gestures |
Step 1 - Concept generation (Day 1): Write 5-10 ad concepts following the hook/body/CTA framework. Use ChatGPT or Claude to brainstorm variations on winning angles. Each concept = 1 hook + 1 narrative + 1 CTA.
Step 2 - Asset creation (Days 1-2):
Step 3 - Assembly (Day 2-3):
Step 4 - Quality filter (Day 3):
Step 5 - Deploy and test (Day 4+):
| Factor | Use AI Creative | Use Traditional Production |
|---|---|---|
| Budget | Under $5K/month creative spend | Over $10K/month creative spend |
| Volume needed | 20+ variations per week | 5-10 polished pieces per month |
| Speed | Need assets in 24-48 hours | Can plan 2-4 week production cycles |
| Product type | Digital, SaaS, info products | Physical products needing real demos |
| Brand stage | Testing/iterating on positioning | Established brand with strict guidelines |
| Languages | Multi-market, 5+ languages | Single market, 1-2 languages |
Hybrid approach (recommended for $5K-$50K/month): AI for 70% of volume (variations, hooks, B-roll, retargeting). Human production for 30% (hero ads, testimonials, demos). Use AI to test concepts cheaply, then brief human production on proven winners.
70% - Proven winners (scale): Allocate to creatives with CPA below target for 5+ days. Increase budget 20-30% every 2-3 days while performance holds. These are your revenue drivers.
20% - Validated iterations (optimize): New variations of winning concepts. Swap hooks on proven bodies. Test new audiences with winning creative. Same messaging, different formats (static to video, video to carousel).
10% - New concepts (test): Completely new angles, formats, and audiences. This is where breakthrough creative comes from. Expect 70-80% of tests to fail. Budget $20-50/day per new concept for 3-5 days minimum.
| Monthly Spend | Winners (70%) | Iterations (20%) | New Tests (10%) | Active Creatives |
|---|---|---|---|---|
| $1K-$3K | $700-$2,100 | $200-$600 | $100-$300 | 5-10 |
| $3K-$10K | $2,100-$7,000 | $600-$2,000 | $300-$1,000 | 10-20 |
| $10K-$30K | $7,000-$21,000 | $2,000-$6,000 | $1,000-$3,000 | 20-40 |
| $30K-$100K | $21,000-$70,000 | $6,000-$20,000 | $3,000-$10,000 | 40-80 |
| $100K+ | $70,000+ | $20,000+ | $10,000+ | 80+ |
| Scenario | Meta | TikTok | YouTube | ||
|---|---|---|---|---|---|
| B2C e-commerce (<$50 AOV) | 40% | 30% | 25% | 5% | 0% |
| B2C e-commerce ($50+ AOV) | 45% | 35% | 15% | 5% | 0% |
| DTC brand launch | 30% | 15% | 40% | 10% | 5% |
| B2B SaaS (SMB) | 35% | 40% | 10% | 10% | 5% |
| B2B SaaS (Enterprise) | 15% | 30% | 5% | 10% | 40% |
| Local service business | 30% | 50% | 10% | 10% | 0% |
| Info products / courses | 35% | 15% | 35% | 15% | 0% |
Allocate 20-30% of total ad spend to creative testing (concept development, AI tools, production).
| Monthly Ad Spend | Creative Test Budget | AI Tool Budget | Production Budget |
|---|---|---|---|
| $1K-$5K | $200-$1,500 | $50-$100 | $150-$1,400 |
| $5K-$15K | $1,000-$4,500 | $100-$300 | $900-$4,200 |
| $15K-$50K | $3,000-$15,000 | $300-$500 | $2,700-$14,500 |
| $50K+ | $10,000+ | $500-$1,000 | $9,500+ |
Build modular creative by mixing and matching components:
3-5 hooks x 2-3 body narratives x 2-3 CTAs = 12-45 unique combinations
Each component is tested independently to isolate what drives performance.
Goal: Find which concepts resonate before optimizing elements. Budget: $20-50/day per concept, test 3-5 concepts simultaneously. Metrics: Hook rate (>25%), CTR (>1%), initial CPA direction. Decision: Advance top 2 concepts to Phase 2. Kill concepts with hook rate <15%.
Goal: Identify winning hooks, bodies, and CTAs independently. Process:
Goal: Scale proven creative while maintaining CPA. Process:
| Week | Focus | Budget/Day | Key Metric |
|---|---|---|---|
| 1 | Launch 3-5 concepts | $100-$250 | Hook rate, CTR |
| 2 | Isolate winning elements | $150-$300 | Hold rate, CPA |
| 3 | Scale 2-3 winners | $200-$500 | CPA, ROAS |
| 4 | Iterate on winners + new concepts | $300-$700 | CPA trend, creative hit rate |
Meta Dynamic Creative: Upload 5 images, 5 headlines, 5 descriptions, 5 CTAs. Meta tests all combinations and optimizes delivery. Review after 50K impressions per element. Best for $10K+/month accounts.
TikTok Smart Creative: Load up to 50 assets per ad. Smart+ auto-selects top performers and recommends new creative from your library. Pair with Symphony for auto-enhancement.
Google PMax Asset Groups: Create themed groups with full asset sets. Google serves the best combination per user and placement. Create 3-5 asset groups per campaign to test themes.
| Signal | Threshold | Action |
|---|---|---|
| CTR declining | 20%+ drop from peak over 3 days | Queue replacement creative |
| CPM rising | 15%+ increase with same targeting | Expand audience or refresh creative |
| Frequency climbing | Above 2.5 for prospecting | Introduce new creative variations |
| Frequency critical | Above 3.0 for retargeting | Immediate creative swap required |
| Hook rate dropping | Below 20% after previously >25% | New hooks needed |
| Comment sentiment | Shift from positive to "I keep seeing this" | Creative is burnt, retire it |
| CPA creep | 15%+ above target for 3+ days | Test new angles, not just variations |
| Platform | Format | Typical Lifespan | Refresh Cadence |
|---|---|---|---|
| TikTok | In-feed video | 1-3 weeks | Weekly new creative |
| TikTok | Spark Ads | 2-4 weeks | Bi-weekly refresh |
| Meta | Reels | 2-3 weeks | Weekly new creative |
| Meta | Feed static | 1-2 weeks | 2x weekly refresh |
| Meta | Carousel | 3-4 weeks | Bi-weekly refresh |
| Google PMax | Mixed assets | 3-6 weeks | Monthly asset refresh |
| YouTube | In-stream | 4-8 weeks | Monthly refresh |
| Sponsored content | 3-5 weeks | Bi-weekly refresh |
Maintain a rolling pipeline to prevent gaps when creative fatigues:
Always have in production: 3-5 concepts in development, 3-5 assets in production, 3-5 in testing queue, 5-10 running live.
Weekly cadence: Mon - review and kill/scale. Tue - brief new creative. Wed - AI generate assets. Thu - refine and assemble. Fri - launch to test campaigns.
| Timeframe | Kill If | Minimum Data Required |
|---|---|---|
| 24 hours | Hook rate < 15% | 5K+ impressions |
| 48 hours | CTR < 0.5% | 2K+ impressions |
| 3 days | CPA > 2x target | $50+ spend or 1K+ clicks |
| 3 days | ROAS < 0.5x target | $100+ spend |
| 5 days | No conversions | 3x target CPA in spend |
| 7 days | CPA trending up 3 consecutive days | Statistically significant data |
| 14 days | CPA 1.5x above target with no improvement | Full test cycle complete |
Never kill before minimum data. Bad decisions from small samples cost more than extra test spend.
| Condition | Action | Frequency |
|---|---|---|
| CPA < target for 48 hours | Increase budget 20% | Every 2-3 days |
| CPA < 50% of target for 72 hours | Increase budget 30-50% | Every 2 days |
| Winner holds after 3 budget increases | Duplicate ad set to new audience | Once per winner |
| Creative at 100K impressions | Commission 3 variations | Immediately |
| Creative at 500K impressions | Test new hooks on same structure | Immediately |
| CPA rises after budget increase | Revert to previous budget, wait 48 hours | As needed |
| CPA holds but volume plateaus | Test broader audiences or new placements | Weekly |
Budget increase limits: Max 30% per day on Meta (algorithm needs recalibration time). Google PMax and TikTok Smart+ tolerate up to 50% increases.
| Platform | Low CPM | Avg CPM | High CPM | Primary Factor |
|---|---|---|---|---|
| TikTok | $3 | $4-$7 | $15 | Audience targeting specificity |
| YouTube | $3 | $6-$8 | $12 | Ad format (bumper vs in-stream) |
| Meta (Reels) | $4 | $7-$10 | $18 | Q4 competition, audience size |
| Meta (Feed) | $5 | $8-$12 | $20 | Industry competition |
| Google Display | $1 | $3-$5 | $10 | Placement quality |
| Google Search | N/A | $10-$30 | $50+ | Keyword competition |
| $20 | $33-$65 | $100+ | Job title targeting specificity |
| Platform | Avg CPC | Range |
|---|---|---|
| TikTok | $0.50-$1.00 | $0.20-$2.00 |
| YouTube | $2.00-$4.00 | $0.50-$6.00 |
| Meta | $0.50-$1.50 | $0.30-$3.00 |
| Google Search | $2.00-$5.00 | $0.50-$50+ |
| Google Display | $0.50-$1.00 | $0.20-$3.00 |
| $5.00-$12.00 | $3.00-$20.00 |
| Channel | Median ROAS | Strong ROAS | Top Performer |
|---|---|---|---|
| Google Search | 4.0-8.0x | 8.0-12.0x | 15x+ |
| Google Shopping | 3.0-6.0x | 6.0-10.0x | 12x+ |
| Meta (overall) | 2.5-4.0x | 4.0-6.0x | 8x+ |
| TikTok | 1.5-3.0x | 3.0-5.0x | 6x+ |
| YouTube | 2.0-4.0x | 4.0-7.0x | 10x+ |
| 1.0-2.0x | 2.0-3.5x | 5x+ | |
| Overall median | 2.19x | 2.87x+ | Industry-dependent |
| Industry | Meta CPA | Google CPA | TikTok CPA |
|---|---|---|---|
| E-commerce (<$50 AOV) | $12-$25 | $15-$30 | $8-$20 |
| E-commerce ($50+ AOV) | $25-$60 | $30-$75 | $15-$40 |
| SaaS (SMB) | $50-$150 | $40-$120 | $30-$80 |
| SaaS (Enterprise) | $150-$500 | $100-$400 | N/A |
| Lead gen (B2B) | $30-$100 | $25-$80 | $20-$60 |
| Education / courses | $20-$60 | $15-$50 | $10-$35 |
| Health / wellness | $15-$40 | $20-$50 | $10-$30 |
| Finance / fintech | $40-$150 | $30-$100 | $25-$80 |
One concept, adapted per platform. Never just repost the same file everywhere.
| Element | TikTok | Meta Reels | Meta Feed | YouTube | |
|---|---|---|---|---|---|
| Tone | Raw, native, trend-aware | Polished casual | Professional casual | Educational, longer | Thought leadership |
| Length | 15-30s | 15-30s | 15-45s (video), N/A (static) | 30s-2min | 15-30s (video), N/A (text) |
| Hook style | Pattern interrupt, trending sound | Visual hook, text overlay | Bold headline, eye-catching image | Question or stat | Insight or contrarian take |
| CTA | "Link in bio" / "Comment X" | "Tap to learn more" | "Shop now" / "Learn more" | "Subscribe" / "Link below" | "Comment your take" / "DM me" |
| Captions | Required, native style | Required, clean | Optional for static | Recommended | Not typical |
| Music/sound | Trending sounds critical | Licensed music | Optional | Background only | Rarely |
| Format | 9:16 vertical | 9:16 vertical | 1:1 or 4:5 | 16:9 or 9:16 | 1:1 or 16:9 |
Level 1 - Platform reporting (baseline): Use each platform's native attribution (Meta Ads Manager, Google Ads, TikTok Ads Manager). Good for directional optimization but overreports due to overlap.
Level 2 - Unified analytics (recommended): Tools like Triple Whale, Cometly, or Rockerbox for cross-platform creative-level attribution. Track which specific creatives drive purchases, not just clicks.
Level 3 - Incrementality testing (advanced): Geo-lift tests and holdout experiments to measure true incremental impact. Tools: Measured, Triple Whale, or platform-native lift studies. Run quarterly on top-spending channels.
Level 4 - Marketing mix modeling (enterprise): Statistical models combining all spend, creative, and revenue data. Best for $100K+/month accounts. Tools: Measured, Recast, internal data science.
| Metric | What It Tells You | Benchmark |
|---|---|---|
| Hook rate | First 3 seconds holding attention | >25% good, >35% great |
| Hold rate | Viewers past 50% of video | >15% |
| Thumb-stop rate | Stopping scroll on static ads | >3% |
| CTR | Interest in clicking through | >1% good, >2% great |
| Conversion rate | Clicks turning into actions | 2-5% landing page |
| Creative win rate | % of tested creatives that beat CPA target | 15-25% is healthy |
| Creative velocity | New creatives launched per week | 5-10 for $10K+/month |
| Time to fatigue | Days until creative performance degrades 20% | Track and increase over time |
Before launching any AI-generated ad:
For checklists, benchmarks, and discovery questions read references/quick-reference.md when you need detailed reference.
Published skills whose descriptions closely match this one, with their headline benchmark lift.
Skills that do similar work, with their headline benchmark lift.
From the same source and nearby in the catalogue.