Ready to supercharge your AI agents?

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

Education Data Explorer · brycewang-stanford | DecimalAI

Databy brycewang-stanford

Education Data Explorer

Discovers education data from Urban Institute Portal: endpoints, variables, year coverage, join keys (CCD, IPEDS, CRDC, Scorecard, SAIPE). Use to map questions to data. Load before education-data-query — discovery here, download there.

Never made a case worse

SkillScore

SkillScore87/100

Test results 87

Safety

3/3 safety checks passed

Code scan
Intent
Content

What it changes

35%87%of test cases passed +52 pts

Without skill8 of 23

With skill20 of 23

12 fixed8 passed either way3 still failing0 broke

Measured on gemini-3.6-flash8/9/2026v1

Track record

Agent runs: 0; All time

See one task, with and without the skill

One test taskI am querying the metadata API to discover available endpoints for school-level data. Which filter parameter works on the metadata API endpoint to restrict results to school-level datasets?

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 brycewang-stanford-education-data-explorer
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("brycewang-stanford-education-data-explorer")
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 brycewang-stanford-education-data-explorer

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/brycewang-stanford-education-data-explorer/SKILL.md https://app.decimal.ai/s/brycewang-stanford-education-data-explorer/SKILL.md

Copy SKILL.md

Save it in your project, as

.claude/skills/brycewang-stanford-education-data-explorer/SKILL.md

This gets SKILL.md only. The skill's 5 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 brycewang-stanford-education-data-explorer

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("brycewang-stanford-education-data-explorer")
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/brycewang-stanford-education-data-explorer/SKILL.md

Give it to your agent

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

skill = Path(".agents/skills/brycewang-stanford-education-data-explorer/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 5 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 brycewang-stanford-education-data-explorer

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("brycewang-stanford-education-data-explorer")
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/brycewang-stanford-education-data-explorer/SKILL.md

Give it to your agent

from pathlib import Path
from agents import Agent, Runner

skill = Path(".agents/skills/brycewang-stanford-education-data-explorer/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 5 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 brycewang-stanford-education-data-explorer

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("brycewang-stanford-education-data-explorer")
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/brycewang-stanford-education-data-explorer/SKILL.md

Give it to your agent

from pathlib import Path
from openai import OpenAI

client = OpenAI()
skill = Path(".agents/skills/brycewang-stanford-education-data-explorer/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 5 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 brycewang-stanford-education-data-explorer

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("brycewang-stanford-education-data-explorer")
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/brycewang-stanford-education-data-explorer/SKILL.md

Give it to your agent

from pathlib import Path
import anthropic

client = anthropic.Anthropic()
skill = Path(".agents/skills/brycewang-stanford-education-data-explorer/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 5 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 brycewang-stanford-education-data-explorer

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("brycewang-stanford-education-data-explorer")
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/brycewang-stanford-education-data-explorer/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/brycewang-stanford-education-data-explorer/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 5 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 brycewang-stanford-education-data-explorer

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("brycewang-stanford-education-data-explorer")
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/brycewang-stanford-education-data-explorer/SKILL.md

Give it to your agent

from pathlib import Path
from langchain.agents import create_agent

skill = Path(".agents/skills/brycewang-stanford-education-data-explorer/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 5 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 brycewang-stanford-education-data-explorer

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("brycewang-stanford-education-data-explorer")
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/brycewang-stanford-education-data-explorer/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 5 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 brycewang-stanford-education-data-explorer

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("brycewang-stanford-education-data-explorer")
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/brycewang-stanford-education-data-explorer/SKILL.md

Give it to your agent

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

skill = Path(".agents/skills/brycewang-stanford-education-data-explorer/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 5 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 brycewang-stanford-education-data-explorer

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("brycewang-stanford-education-data-explorer")
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/brycewang-stanford-education-data-explorer/SKILL.md

Give it to your agent

from pathlib import Path
from pydantic_ai import Agent

skill = Path(".agents/skills/brycewang-stanford-education-data-explorer/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 5 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 brycewang-stanford-education-data-explorer

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("brycewang-stanford-education-data-explorer")
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/brycewang-stanford-education-data-explorer/SKILL.md

Give it to your agent

from pathlib import Path
from deepagents import create_deep_agent

skill = Path(".agents/skills/brycewang-stanford-education-data-explorer/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 5 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 brycewang-stanford-education-data-explorer

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("brycewang-stanford-education-data-explorer")
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/brycewang-stanford-education-data-explorer/SKILL.md

Give it to your agent

import asyncio
from pathlib import Path
from agent_framework import Agent

skill = Path(".agents/skills/brycewang-stanford-education-data-explorer/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 5 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 brycewang-stanford-education-data-explorer

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("brycewang-stanford-education-data-explorer")
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/brycewang-stanford-education-data-explorer/SKILL.md

Give it to your agent

from pathlib import Path
from strands import Agent

skill = Path(".agents/skills/brycewang-stanford-education-data-explorer/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 5 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 brycewang-stanford-education-data-explorer

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/brycewang-stanford-education-data-explorer/SKILL.md https://app.decimal.ai/s/brycewang-stanford-education-data-explorer/SKILL.md

Copy SKILL.md

Save it in your project, as

.agents/skills/brycewang-stanford-education-data-explorer/SKILL.md

This gets SKILL.md only. The skill's 5 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/brycewang-stanford-education-data-explorer/SKILL.md

Read the summary as JSON

curl -fsSL https://app.decimal.ai/s/brycewang-stanford-education-data-explorer.json

This gets SKILL.md only. The skill's 5 other files are not included.

Other ways to add it

Other setups docs

You don't need a DecimalAI account.

Version: v1, tested 8/9/2026
Source: brycewang-stanford/Auto-Empirical-Research-Skills ↗
License: NOASSERTION

Manage it in DecimalAI

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

Use this skill

SKILL.md

Education Data Explorer

Discovers available education data from the Urban Institute Education Data Portal: endpoints, variables, year coverage, and join keys for schools, districts, and colleges (CCD, IPEDS, CRDC, Scorecard, SAIPE, and more). Use during discovery and scoping phases when identifying what data exists, mapping research questions to endpoints, or resolving variable name discrepancies between documentation and actual field names. Load before education-data-query — this skill covers discovery; education-data-query handles the download.

Discover available education data from the Urban Institute Education Data Portal for research planning and query design.

What is the Education Data Portal?

Comprehensive education data from Urban Institute - free and publicly available
Three data levels: schools, school-districts, college-university
Multiple data sources: CCD, IPEDS, CRDC, College Scorecard, EDFacts, SAIPE, FSA, MEPS, PSEO, etc.
Coverage: 1980-2023 depending on source
Access: Mirror downloads (parquet/CSV) via education-data-query skill
Documentation: https://educationdata.urban.org/documentation/
Curation layer, not a direct pass-through: The EDP standardizes data from original federal sources with lowercase variable names, integer-encoded categoricals, and uniform missing value codes (-1, -2, -3)
Coverage varies by source: Some sources are fully mirrored; others are partially mirrored with only a subset of variables or datasets available. See individual education-data-source-* skills for source-specific coverage details

Skill Provenance Note: Each *-data-source-* skill includes provenance.skill_last_updated in its frontmatter. When exploring data sources during Stage 2, note the provenance dates of any skills you reference — if more than a few months old, flag this in your findings so the orchestrator can consider re-verifying with data-ingest.

Note: This workflow uses mirror-based file downloads, not paginated API calls. See education-data-query skill for fetch patterns and datasets-reference.md for file paths.

Reference File Structure

File	Purpose	When to Read
`schools-endpoints.md`	All school-level endpoints and variables	Researching K-12 schools
`districts-endpoints.md`	All district-level endpoints and variables	Researching school districts
`colleges-endpoints.md`	All college-level endpoints and variables	Researching higher education
`variable-codes.md`	Code values for states, grades, race, etc.	Interpreting or filtering data
`metadata-api.md`	Programmatic endpoint/variable discovery	Dynamic exploration

Decision Trees

What data level do I need?

What entity am I researching?
├─ Individual K-12 schools → schools level
│   └─ See ./references/schools-endpoints.md
├─ School districts / LEAs → school-districts level
│   └─ See ./references/districts-endpoints.md
├─ Colleges / Universities → college-university level
│   └─ See ./references/colleges-endpoints.md
└─ Not sure
    ├─ Need school-specific data (discipline, AP, demographics) → schools
    ├─ Need aggregate district data (finance, poverty) → school-districts
    └─ Need postsecondary data (enrollment, aid, outcomes) → college-university

What topic am I researching?

Research topic?
├─ Enrollment / Demographics
│   ├─ K-12 public schools → CCD enrollment endpoints
│   ├─ Civil rights indicators → CRDC enrollment
│   └─ Colleges → IPEDS enrollment
├─ School Finance
│   ├─ District revenue/expenditure → CCD finance
│   └─ College finance → IPEDS finance
├─ Student Outcomes
│   ├─ K-12 assessments → EDFacts
│   ├─ Graduation rates (K-12) → EDFacts
│   ├─ College completion → IPEDS completions
│   └─ Post-college earnings → College Scorecard
├─ Student Aid / Loans
│   ├─ College financial aid → IPEDS aid
│   ├─ Federal loans/grants → FSA
│   └─ Debt/repayment → College Scorecard
├─ Discipline / Civil Rights
│   └─ K-12 discipline, harassment, restraint → CRDC
├─ Poverty Estimates
│   └─ District-level → SAIPE
└─ Directory / Location
    ├─ K-12 schools → CCD directory
    ├─ Districts → CCD directory
    └─ Colleges → IPEDS directory

How do I find specific variables?

Finding variables?
├─ Know the endpoint → Check reference file for variable list
├─ Know the topic → Use topic index below
├─ Need to search programmatically → See ./references/metadata-api.md
└─ Need code definitions → See ./references/variable-codes.md

Quick Reference: Data Levels

Level	Key Sources	Primary ID	ID Format
schools	CCD, CRDC, EDFacts, MEPS, NHGIS	`ncessch`	12-char string
school-districts	CCD, SAIPE, EDFacts	`leaid`	7-char string
college-university	IPEDS, Scorecard, FSA, PSEO, EADA	`unitid`	6-digit integer

Quick Reference: Data Sources

Source	Level	Description	Years
CCD	Schools, Districts	Public K-12 directory, enrollment, finance	1986-2023
CRDC	Schools	Civil rights indicators, discipline, AP courses	2011-2021
EDFacts	Schools, Districts	Assessments, graduation rates	2009-2020
IPEDS	Colleges	Enrollment, completions, finance, institutional data	1980-2023
College Scorecard	Colleges	Earnings, debt, student outcomes	1996-2020
SAIPE	Districts	Census poverty estimates for school-age children	1995-2023
FSA	Colleges	Federal student aid, loans, grants, 90/10	1999-2021
MEPS	Schools	School poverty measure	2006-2019
NHGIS	Schools	Census geography crosswalks	1990, 2000, 2010, 2020

Quick Reference: Common Endpoints

Schools

Endpoint	Description
`/schools/ccd/directory/{year}/`	School directory (location, type, enrollment)
`/schools/ccd/enrollment/{year}/{grade}/`	Enrollment by grade
`/schools/crdc/discipline/{year}/`	Discipline incidents
`/schools/crdc/ap-ib-enrollment/{year}/race/sex/`	AP/IB enrollment (requires disaggregation)
`/schools/edfacts/assessments/{year}/{grade}/`	Assessment results

Districts

Endpoint	Description
`/school-districts/ccd/directory/{year}/`	District directory
`/school-districts/ccd/enrollment/{year}/{grade}/`	District enrollment
`/school-districts/ccd/finance/{year}/`	Revenue and expenditure
`/school-districts/saipe/{year}/`	Poverty estimates

Colleges

Endpoint	Description
`/college-university/ipeds/directory/{year}/`	Institution directory
`/college-university/ipeds/admissions-enrollment/{year}/`	Admissions data
`/college-university/ipeds/enrollment-full-time-equivalent/{year}/`	FTE enrollment
`/college-university/ipeds/fall-enrollment/{year}/{level}/`	Fall enrollment
`/college-university/ipeds/graduation-rates/{year}/`	Graduation rates
`/college-university/scorecard/earnings/{year}/`	Post-college earnings

Exploration Workflow

Follow these steps to identify data for a research question:

Identify data level
- Schools: individual K-12 school records
- Districts: school district / LEA records
- Colleges: postsecondary institution records
Identify relevant data source(s)
- Use the data sources table above
- Multiple sources may be needed (e.g., CCD + CRDC)
Check available endpoints
- Read the appropriate reference file
- Note endpoint URL pattern and variables
Review variables and filters
- Check variable lists in reference files
- Note which variables can be used as filters
Check years available
- Each endpoint has different year coverage
- Use metadata API to get exact years
Understand source context (RECOMMENDED)
- Load the appropriate education-data-source-* skill for deep context
- Understand data collection methodology and limitations
- Review variable definitions and coding schemes
Plan query
- Load education-data-query skill for query construction
- Or use metadata API to build query programmatically

URL Pattern Structure

All endpoints follow this pattern:

/api/v1/{level}/{source}/{topic}/{year}/[{disaggregation}/]

Examples:

/api/v1/schools/ccd/directory/2022/
/api/v1/schools/ccd/enrollment/2022/grade-5/
/api/v1/schools/ccd/enrollment/2022/grade-5/race/
/api/v1/school-districts/ccd/finance/2021/
/api/v1/college-university/ipeds/fall-enrollment/2022/undergraduate/

Filtering

Query Parameters

Parameter	Description	Example
`fips`	State FIPS code	`?fips=6` (California)
`leaid`	District ID	`?leaid=0600001`
`ncessch`	School ID	`?ncessch=060000100001`
`unitid`	College ID	`?unitid=110635`
`year`	Filter by year	`?year=2022`

Response Format

{
  "count": 12345,
  "next": "https://educationdata.urban.org/api/v1/...?page=2",
  "previous": null,
  "results": [
    {"ncessch": "...", "school_name": "...", ...},
    ...
  ]
}

Cross-Reference to Related Skills

Skill	Purpose	When to Use
`education-data-query`	Download data from mirrors	After identifying endpoints/variables
`education-data-context`	Interpret data, understand limitations	After retrieving data

Deep-Dive Data Source Skills

For comprehensive understanding of each data source beyond the portal documentation, load the appropriate source-specific skill:

Skill	Data Source	Key Topics
`education-data-source-ccd`	Common Core of Data	K-12 directory, enrollment, finance, staffing surveys
`education-data-source-crdc`	Civil Rights Data Collection	Discipline, harassment, course access, civil rights context
`education-data-source-saipe`	Small Area Income & Poverty	District poverty estimates, model methodology
`education-data-source-edfacts`	EDFacts	State assessments, graduation rates, accountability
`education-data-source-ipeds`	IPEDS	College enrollment, graduation, finance, completions
`education-data-source-scorecard`	College Scorecard	Post-college earnings, debt, repayment
`education-data-source-nhgis`	NHGIS	Census geography, demographic crosswalks
`education-data-source-fsa`	Federal Student Aid	Pell, loans, financial responsibility, 90/10
`education-data-source-nacubo`	NACUBO	College endowment data
`education-data-source-nccs`	NCCS	Nonprofit data for private colleges
`education-data-source-meps`	MEPS	Model-based school poverty (superior to FRPL)
`education-data-source-eada`	EADA	College athletics equity data
`education-data-source-campus-safety`	Campus Safety	Campus crime, Clery Act data
`education-data-source-pseo`	PSEO	Post-graduation employment outcomes

When to load source skills:

Need deeper understanding of data collection methodology
Encountering unexpected values or patterns
Planning analysis that requires understanding source limitations
Working with less common data elements not covered in this skill

Topic Index

Topic	Reference File	Section
School directory	`schools-endpoints.md`	CCD Directory
School enrollment	`schools-endpoints.md`	CCD Enrollment
Discipline data	`schools-endpoints.md`	CRDC Discipline
AP/IB courses	`schools-endpoints.md`	CRDC AP-IB-GT
K-12 assessments	`schools-endpoints.md`	EDFacts
District directory	`districts-endpoints.md`	CCD Directory
District finance	`districts-endpoints.md`	CCD Finance
District poverty	`districts-endpoints.md`	SAIPE
College directory	`colleges-endpoints.md`	IPEDS Directory
College enrollment	`colleges-endpoints.md`	IPEDS Enrollment
College graduation	`colleges-endpoints.md`	IPEDS Graduation
Financial aid	`colleges-endpoints.md`	IPEDS Aid, FSA
Post-college earnings	`colleges-endpoints.md`	Scorecard
Student debt	`colleges-endpoints.md`	Scorecard, FSA
State FIPS codes	`variable-codes.md`	State FIPS
Grade codes	`variable-codes.md`	Grade Codes
Race/ethnicity codes	`variable-codes.md`	Race Codes
Locale codes	`variable-codes.md`	Urban-Centric Locale
Programmatic discovery	`metadata-api.md`	All

Example: Planning a Research Query

Research question: "What is the relationship between school poverty and AP course offerings in California high schools?"

Data level: Schools (individual school records)
Data sources needed:
- CRDC for AP course data
- MEPS or CCD for poverty measure
Endpoints:
- /schools/crdc/ap-ib-enrollment/{year}/race/sex/ - AP enrollment (requires disaggregation)
- /schools/meps/{year}/ - School poverty measure
Key variables:
- ncessch - school identifier (for joining)
- fips=6 - California filter
- AP enrollment variables from CRDC
- Poverty measure from MEPS
Years: Check overlap (CRDC: 2011-2021, MEPS: 2006-2019)
Next step: Load education-data-query skill to construct the actual API calls

Common Pitfalls

Year coverage varies: Always check years available for each endpoint
Different ID formats: ncessch (12-char), leaid (7-char), unitid (6-digit)
Disaggregation in URL: Grade, race, sex are often URL path components, not query params
Missing data codes: -1, -2, -3 have specific meanings (see variable-codes.md)

Pre-Query Validation

CRITICAL: Variable Name Discrepancies

The Education Data Portal API often uses different variable names than documentation suggests. Always fetch a sample first:

# Test query to verify actual column names
response = requests.get(
    "https://educationdata.urban.org/api/v1/college-university/ipeds/directory/2023/"
)
data = response.json()
print("Actual columns:", list(data['results'][0].keys()))

Known discrepancies:

Documented	Actual API Field	Endpoint
`inst_level`	`institution_level`	IPEDS Directory
`applicants_total`	`number_applied`	IPEDS Admissions
`admissions_total`	`number_admitted`	IPEDS Admissions
`grad_rate_150pct`	`completion_rate_150pct`	IPEDS Graduation Rates
`school_poverty`	`meps_poverty_pct`	MEPS
`population_5_17_poverty`	`est_population_5_17_poverty`	SAIPE

See the relevant education-data-source-* skill for comprehensive variable mappings per source.

Metadata API Limitations

The metadata API has undocumented limitations:

?section=schools works to filter by data level
?source=ipeds does NOT work - filter client-side instead
Response field names differ: source is actually class_name, source_name is actually label

Data Source Details

Quick summaries below. For comprehensive documentation including methodology, variable definitions, data quality issues, and historical changes, load the corresponding education-data-source-* skill.

CCD (Common Core of Data)

Coverage: All public elementary and secondary schools and districts in the U.S.

Topic	Schools	Districts
Directory	Yes	Yes
Enrollment	Yes (by grade, race, sex)	Yes (by grade, race, sex)
Finance	No	Yes (revenue, expenditure)

Key Variables:

ncessch: 12-character NCES school ID
leaid: 7-character NCES district ID
enrollment: Total enrollment count
free_or_reduced_price_lunch: FRPL-eligible students (poverty proxy)
charter: Charter school indicator
urban_centric_locale: Urban/suburban/town/rural classification

Deep dive: Load education-data-source-ccd for survey components, data collection process, variable coding, state variations, and historical changes (e.g., 2006 locale code revision, 2010 race category changes).

CRDC (Civil Rights Data Collection)

Coverage: Biennial survey of public schools (2011, 2013, 2015, 2017, 2020, 2021)

Topics:

Discipline (suspensions, expulsions, arrests)
Chronic absenteeism
Harassment and bullying
Restraint and seclusion
Advanced courses (AP, IB, gifted)
Course offerings
Teacher qualifications
Retention
COVID impacts (2020 only)

Key Feature: Disaggregation by race, sex, disability, and LEP status

Deep dive: Load education-data-source-crdc for civil rights legal context (Title VI, IX, Section 504), collection methodology, underreporting issues, and year-to-year changes.

EDFacts

Coverage: State assessment and accountability data

Topics:

Assessment proficiency rates (reading, math)
Graduation rates (4-year adjusted cohort)

Key Feature: Data available by special populations (disability, economically disadvantaged, LEP, homeless, migrant, foster care)

CRITICAL: State assessment scores CANNOT be compared across states (different tests, cut scores).

Deep dive: Load education-data-source-edfacts for ESSA/NCLB accountability context, why cross-state comparison is invalid, ACGR methodology, and subgroup reporting rules.

IPEDS (Integrated Postsecondary Education Data System)

Coverage: All Title IV-eligible postsecondary institutions

Topics:

Institutional characteristics and directory
Admissions and enrollment
Student charges (tuition, fees, room, board)
Financial aid
Finance (revenue, expenditure, assets)
Graduation rates
Completions (degrees awarded by CIP code)
Human resources (salaries, faculty)

Key Variables:

unitid: 6-digit IPEDS institution ID
inst_control: 1=Public, 2=Private nonprofit, 3=Private for-profit
institution_level: 1=Less-than-2-year, 2=2-year, 4=4-year (no code 3)
hbcu: Historically Black college indicator

Deep dive: Load education-data-source-ipeds for critical graduation rate limitations (first-time full-time only), GASB vs FASB finance accounting, survey components, and identifier changes.

College Scorecard

Coverage: Title IV institutions with outcome data

Topics:

Post-college earnings (6 and 10 years after entry)
Student debt and repayment
Default rates
Completion rates by income level

Key Feature: Links education to labor market outcomes

CRITICAL: Only covers Title IV aid recipients (selection bias toward lower-income students).

Deep dive: Load education-data-source-scorecard for earnings methodology (IRS data), population coverage limitations, suppression rules, and field-of-study data.

SAIPE (Small Area Income and Poverty Estimates)

Coverage: Census Bureau poverty estimates for school districts

Key Variables:

population_5_17_poverty: Children 5-17 in poverty
population_5_17_poverty_pct: Percent in poverty
median_household_income: District median income

Deep dive: Load education-data-source-saipe for model-based estimation methodology, confidence intervals (not available at district level), and comparison to other poverty measures.

FSA (Federal Student Aid)

Coverage: Title IV institutions receiving federal aid

Topics:

Pell grants
Direct loans (subsidized, unsubsidized, PLUS)
Campus-based aid (Perkins, work-study)
Financial responsibility scores
90/10 revenue (for-profit institutions)

Deep dive: Load education-data-source-fsa for Title IV program details, financial responsibility composite scores, and 90/10 rule compliance.

Additional Data Sources

Source	Coverage	Deep Dive Skill
MEPS	School-level poverty estimates (superior to FRPL for cross-state comparison)	`education-data-source-meps`
NHGIS	Census geography crosswalks for schools	`education-data-source-nhgis`
NACUBO	College endowment data	`education-data-source-nacubo`
NCCS	Nonprofit data for private colleges (Form 990)	`education-data-source-nccs`
EADA	College athletics equity data	`education-data-source-eada`
Campus Safety	Campus crime statistics (Clery Act)	`education-data-source-campus-safety`
PSEO	Post-graduation employment outcomes	`education-data-source-pseo`

Joining Data Across Sources

School-Level Joins

Join school data across sources using ncessch:

Source 1	Source 2	Join Key	Use Case
CCD	CRDC	`ncessch`	Enrollment + discipline
CCD	EDFacts	`ncessch`	Directory + assessments
CCD	MEPS	`ncessch`	Enrollment + poverty
CRDC	MEPS	`ncessch`	AP courses + poverty

Note: Match on year when joining (years may not align perfectly)

District-Level Joins

Join district data using leaid:

Source 1	Source 2	Join Key	Use Case
CCD Directory	CCD Finance	`leaid`	Characteristics + spending
CCD	SAIPE	`leaid`	Enrollment + poverty
CCD	EDFacts	`leaid`	Enrollment + outcomes

College-Level Joins

Join college data using unitid:

Source 1	Source 2	Join Key	Use Case
IPEDS Directory	IPEDS Finance	`unitid`	Characteristics + finance
IPEDS	Scorecard	`unitid`	Enrollment + earnings
IPEDS	FSA	`unitid`	Enrollment + aid data

Disaggregation Patterns

URL Path Disaggregation

Some disaggregations are part of the URL path:

/schools/ccd/enrollment/{year}/{grade}/           # By grade
/schools/ccd/enrollment/{year}/{grade}/race/      # By grade and race
/schools/ccd/enrollment/{year}/{grade}/race/sex/  # By grade, race, and sex

Query Parameter Disaggregation

Other filters are query parameters:

?fips=6                    # California only
?charter=1                 # Charter schools only
?school_level=3            # High schools only
?urban_centric_locale=11   # Large cities only

Available Disaggregations by Source

Source	Grade	Race	Sex	Disability	Econ Status	LEP
CCD	Yes	Yes	Yes	No	No	No
CRDC	No	Yes	Yes	Yes	No	Yes
EDFacts	Yes	Yes	Yes	Yes	Yes	Yes
IPEDS	Level	Yes	Yes	No	No	No

Year Coverage Quick Reference

Source	Earliest	Latest	Update Frequency
CCD Directory	1986	2023	Annual
CCD Finance	1989	2021	Annual (2-year lag)
CRDC	2011	2021	Biennial
EDFacts	2009	2020	Annual
IPEDS	1980	2023	Annual
Scorecard	1996	2020	Annual
SAIPE	1995	2023	Annual
FSA	1999	2021	Annual

Example Research Scenarios

Scenario	Data Sources	Key Variables
Charter vs traditional school outcomes	CCD directory + EDFacts assessments	`charter`, `read_test_pct_prof_midpt`
College affordability by income	IPEDS directory + net-price-by-income	`inst_control`, `income_level`, `avg_net_price`
Discipline disparities by race	CRDC discipline + enrollment (by race)	`race`, `oss_one`, `expulsions_*`
Spending and graduation rates	CCD finance + EDFacts grad-rates	`exp_current_per_pupil`, `grad_rate_midpt`
School poverty and AP access	CRDC ap-ib-enrollment + MEPS	`ap_enrollment`, `meps_poverty_pct`
College earnings by major	IPEDS completions + Scorecard earnings	`cip_code`, `earn_median_wne_p10`

About this skill

Automatic

Data Analyst pack

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/2/2026

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

brycewang-stanford / Auto-Empirical-Research-Skills

skills/17-DAAF-Contribution-Community-daaf/dot-claude/skills/education-data-explorer/SKILL.md

View on GitHub ↗

License NOASSERTIONImported 8/2/2026Version v1Bundled 5 files

Illustrative Want to see what brycewang-stanford/education-data-explorer 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

Education Data Explorer

What is the Education Data Portal?

Reference File Structure

Decision Trees

What data level do I need?

What topic am I researching?

How do I find specific variables?

Quick Reference: Data Levels

Quick Reference: Data Sources

Quick Reference: Common Endpoints

Schools

Districts

Colleges

Exploration Workflow

URL Pattern Structure

Filtering

Query Parameters

Response Format

Cross-Reference to Related Skills

Deep-Dive Data Source Skills

Topic Index

Example: Planning a Research Query

Common Pitfalls

Pre-Query Validation

CRITICAL: Variable Name Discrepancies

Metadata API Limitations

Data Source Details

CCD (Common Core of Data)

CRDC (Civil Rights Data Collection)

EDFacts

IPEDS (Integrated Postsecondary Education Data System)

College Scorecard

SAIPE (Small Area Income and Poverty Estimates)

FSA (Federal Student Aid)

Additional Data Sources

Joining Data Across Sources

School-Level Joins

District-Level Joins

College-Level Joins

Disaggregation Patterns

URL Path Disaggregation

Query Parameter Disaggregation

Available Disaggregations by Source

Year Coverage Quick Reference

Example Research Scenarios

SkillSafety

Source

Related skills

More in the registry