Ready to supercharge your AI agents?

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

Biological Sequence Analysis · mims-harvard | DecimalAI

Scienceby mims-harvard

Biological Sequence Analysis

Biological sequence analysis — gene/protein sequence retrieval (NCBI, Ensembl, UniProt), nucleotide/protein search, ortholog discovery, and FASTQ QC + alignment workflows (Trimmomatic, BWA, samtools, coverage depth).

Maintained

Safety

3/3 safety checks passed

Code scan
Intent
Content

What it changes

100%0%of checks passed -100 pts

Without skill100%

With skill0%

0 fixed1 passed either way22 still failing0 broke

Measured on gemini-3.6-flash7/27/2026v1 · latest is v218 comparable

This benchmark measured an earlier version of the skill.

Track record

Agent runs: 0; All time

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 tooluniverse-sequence-analysis
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("tooluniverse-sequence-analysis")
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 tooluniverse-sequence-analysis

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/tooluniverse-sequence-analysis/SKILL.md https://app.decimal.ai/s/tooluniverse-sequence-analysis/SKILL.md

Copy SKILL.md

Save it in your project, as

.claude/skills/tooluniverse-sequence-analysis/SKILL.md

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

Other ways to add it

Claude Code docs

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

Install

pip install "decimalai[claude-agent-sdk]"
decimalai skills add tooluniverse-sequence-analysis

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("tooluniverse-sequence-analysis")
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/tooluniverse-sequence-analysis/SKILL.md

Give it to your agent

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

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

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

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

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

Other ways to add it

Claude Agent SDK docs

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

Install

pip install "decimalai[openai-agents]"
decimalai skills add tooluniverse-sequence-analysis

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("tooluniverse-sequence-analysis")
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/tooluniverse-sequence-analysis/SKILL.md

Give it to your agent

from pathlib import Path
from agents import Agent, Runner

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

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

Other ways to add it

OpenAI Agents docs

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

Install

pip install "decimalai[openai]"
decimalai skills add tooluniverse-sequence-analysis

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("tooluniverse-sequence-analysis")
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/tooluniverse-sequence-analysis/SKILL.md

Give it to your agent

from pathlib import Path
from openai import OpenAI

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

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

Other ways to add it

OpenAI SDK docs

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

Install

pip install "decimalai[anthropic]"
decimalai skills add tooluniverse-sequence-analysis

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("tooluniverse-sequence-analysis")
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/tooluniverse-sequence-analysis/SKILL.md

Give it to your agent

from pathlib import Path
import anthropic

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

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

Other ways to add it

Anthropic docs

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

Install

pip install "decimalai[google-genai]"
decimalai skills add tooluniverse-sequence-analysis

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("tooluniverse-sequence-analysis")
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/tooluniverse-sequence-analysis/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/tooluniverse-sequence-analysis/SKILL.md").read_text()
response = client.models.generate_content(
    model="gemini-3.5-flash",
    contents="Hello!",
    config=types.GenerateContentConfig(system_instruction=f"You are a helpful agent.\n\n{skill}"),
)

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

Other ways to add it

Gemini API docs

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

Install

pip install "decimalai[langchain]" langchain-openai
decimalai skills add tooluniverse-sequence-analysis

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("tooluniverse-sequence-analysis")
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/tooluniverse-sequence-analysis/SKILL.md

Give it to your agent

from pathlib import Path
from langchain.agents import create_agent

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

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

Other ways to add it

LangChain docs

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

Install

pip install "decimalai>=0.16" "crewai>=1.15.27"
decimalai skills add tooluniverse-sequence-analysis

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("tooluniverse-sequence-analysis")
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/tooluniverse-sequence-analysis/SKILL.md

Give it to your agent

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

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

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

Other ways to add it

CrewAI docs

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

Install

pip install "decimalai[adk]"
decimalai skills add tooluniverse-sequence-analysis

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("tooluniverse-sequence-analysis")
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/tooluniverse-sequence-analysis/SKILL.md

Give it to your agent

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

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

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

Other ways to add it

Google ADK docs

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

Install

pip install "decimalai[pydantic-ai]"
decimalai skills add tooluniverse-sequence-analysis

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("tooluniverse-sequence-analysis")
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/tooluniverse-sequence-analysis/SKILL.md

Give it to your agent

from pathlib import Path
from pydantic_ai import Agent

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

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

Other ways to add it

Pydantic AI docs

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

Install

pip install "decimalai>=0.16" deepagents
decimalai skills add tooluniverse-sequence-analysis

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("tooluniverse-sequence-analysis")
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/tooluniverse-sequence-analysis/SKILL.md

Give it to your agent

from pathlib import Path
from deepagents import create_deep_agent

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

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

Other ways to add it

Deep Agents docs

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

Install

pip install "decimalai>=0.16" agent-framework
decimalai skills add tooluniverse-sequence-analysis

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("tooluniverse-sequence-analysis")
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/tooluniverse-sequence-analysis/SKILL.md

Give it to your agent

import asyncio
from pathlib import Path
from agent_framework import Agent

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

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

Other ways to add it

Microsoft Agent Framework docs

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

Install

pip install "decimalai>=0.16" strands-agents
decimalai skills add tooluniverse-sequence-analysis

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("tooluniverse-sequence-analysis")
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/tooluniverse-sequence-analysis/SKILL.md

Give it to your agent

from pathlib import Path
from strands import Agent

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

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

Other ways to add it

Strands docs

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

pip install "decimalai>=0.15"
decimalai skills add tooluniverse-sequence-analysis

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/tooluniverse-sequence-analysis/SKILL.md https://app.decimal.ai/s/tooluniverse-sequence-analysis/SKILL.md

Copy SKILL.md

Save it in your project, as

.agents/skills/tooluniverse-sequence-analysis/SKILL.md

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

Other ways to add it

Cursor and other editors docs

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

Get the file

curl -fsSL https://app.decimal.ai/s/tooluniverse-sequence-analysis/SKILL.md

Read the summary as JSON

curl -fsSL https://app.decimal.ai/s/tooluniverse-sequence-analysis.json

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

Other ways to add it

Other setups docs

You don't need a DecimalAI account.

Version: v2 (v1 tested 7/27/2026)
Source: mims-harvard/ToolUniverse ↗
License: Apache-2.0

Manage it in DecimalAI

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

Use this skill

SKILL.md

Biological Sequence Analysis

⚠️ TOP-OF-MIND RULE: Trimmomatic "reads completely discarded" = `F + R + 2*D`, summed across samples

When a question asks about Trimmomatic "reads completely discarded" / "reads thrown out" / "reads not in any output", do NOT report the Dropped field alone. Dropped counts PAIRS where both mates failed; each pair = 2 individual reads. Plus the Forward-only and Reverse-only buckets also discard one read per pair.

reads_discarded = sum over samples of (Forward_Only + Reverse_Only + 2 * Dropped)

❌ WRONG: sum(Dropped per sample) — typically reports ~thousands, GT is 100×+ higher

✅ RIGHT: sum(F + R + 2*D per sample)

Full counter table is in the FASTQ section below.

RULE ZERO — Check for pre-computed results FIRST

Before following any instruction below, scan the data folder for:

*_executed.ipynb → read with tu run read_executed_notebook '{"data_folder":"<path>","search":"<keyword>"}' and cite its cell outputs as the authoritative answer
Pre-computed result files (CSV/TSV with names like *results*, *deseq*, *enrich*, *stats*, *_simplified.csv) → read directly and report the requested value
Canonical analysis scripts (analysis.R, run_*.py, find_*.R, *.Rmd) → execute as-is and read the output

Only follow this skill's re-analysis recipe below if none of the above exist. Re-running from raw data produces different numbers than the published answer and is much slower (often 5-10× turn count).

Retrieve, annotate, and compare biological sequences from NCBI, Ensembl, and UniProt. Covers nucleotide search, sequence fetching, gene summaries, ortholog discovery, and protein sequence extraction.

FASTQ QC, Trimmomatic, and read alignment (when raw reads are present)

When the data folder has *.fastq files and the question involves Trimmomatic, BWA, samtools, FastQC, or coverage depth, this skill is the entry point — but the actual work is shell-level (no specific ToolUniverse data tool).

Trimmomatic PE — counting "completely discarded" reads

Trimmomatic PE classifies each input pair as:

Both Surviving (B): R1 and R2 both pass → kept as paired
Forward Only Surviving (F): R1 passes, R2 dropped → R1 kept as singleton, R2 fully discarded
Reverse Only Surviving (R): R2 passes, R1 dropped → R2 kept as singleton, R1 fully discarded
Dropped (D): both fail → BOTH R1 and R2 fully discarded

Counter selection — read the question carefully. CRITICAL: "READS completely discarded" ≠ "PAIRS dropped". The Trimmomatic Dropped count counts PAIRS (each = 2 individual reads). When the question asks about reads (not pairs), translate every counter to per-read terms:

Question phrasing	Formula (per-sample, then SUM across all samples)
"reads completely discarded", "reads thrown out", "reads not in any output"	`F + R + 2*D` (every individual R1 or R2 not in any output FASTQ)
"read pairs dropped", "pairs where both mates failed"	`D`
"individual R2 reads dropped" (R1 kept as singleton)	`F`
"individual R1 reads dropped" (R2 kept as singleton)	`R`
"reads passing QC" / "surviving reads"	`2*B + F + R`

Trimmomatic's stderr summary gives Input Read Pairs: N Both Surviving: B (b%) Forward Only Surviving: F (f%) Reverse Only Surviving: R (r%) Dropped: D (d%). Always sum across ALL input sample pairs (e.g., SRR1 + SRR2 + ...).

DO NOT report just D as "reads completely discarded" — that's pair count, not read count, and is off by ~100×. The "Forward Only" R2 mate IS discarded; the "Reverse Only" R1 mate IS discarded; "Dropped" pairs lose BOTH reads.

Coverage depth (samtools depth / mosdepth)

For "average coverage depth", run samtools depth -a alignment.bam | awk '{sum+=$3; n++} END {print sum/n}' — the -a flag includes positions with zero coverage (otherwise the average is biased upward). For per-chromosome coverage, group by $1.

When to Use

"Get the mRNA sequence for BRCA1"
"Search NCBI for E. coli K-12 complete genome"
"Find orthologs of TP53 across species"
"Fetch the protein sequence for UniProt P04637"
"Get the CDS sequence for Ensembl transcript ENST00000269305"

Workflow

Input -> Phase 1: Gene ID resolution -> Phase 2: Nucleotide retrieval
      -> Phase 3: Protein sequences -> Phase 4: Orthologs -> Output

Phase 1: Gene Identification and Summary

NCBIGene_search: term (string REQUIRED, format "TP53[Symbol] AND Homo sapiens[Organism]"), retmax (int, default 10). Returns {status, data: {esearchresult: {idlist: ["7157"]}}}.

NCBIGene_get_summary: id (string REQUIRED, e.g., "7157"). Returns {status, data: {result: {"7157": {name, description, summary, chromosome, maplocation, genomicinfo, mim}}}}. Result is keyed by gene ID string.

NCBIDatasets_get_gene_by_symbol: symbol (string REQUIRED, e.g., "BRCA1"), taxon (string, e.g., "human"). Returns gene ID, description, location, cross-references.

NCBIDatasets_get_gene: gene_id (string REQUIRED, e.g., "7157"). Returns comprehensive gene info.

Phase 2: Nucleotide Sequence Search and Retrieval

NCBI_search_nucleotide: query (free-form), organism (string), gene (string), strain (string), keywords (string), seq_type ("complete_genome"/"mRNA"/"refseq"), limit (int, default 20). Returns {status, data: {uids: [...], accessions: [...]}}.

NCBI_fetch_accessions: uids (array REQUIRED, e.g., ["545778205"]). Returns {status, data: ["U00096.3"], count: 1}.

NCBI_get_sequence: accession (string REQUIRED, e.g., "NM_007294"), format ("fasta"/"gb"/"embl"). Returns {status, data: "FASTA string...", accession, format, length}.

EnsemblSeq_get_region_sequence: region (string REQUIRED, "chr:start-end", e.g., "17:7668421-7668520"), species (default "homo_sapiens"). Returns {status, data: {sequence, sequence_length}}.

ensembl_get_sequence: id (string REQUIRED, Ensembl ID), type ("genomic"/"cds"/"cdna"/"protein"), multiple_sequences (bool). Returns sequence data.

Gotchas:

NCBI_search_nucleotide returns UIDs, not accessions. Use NCBI_fetch_accessions to convert.
NCBI_fetch_accessions requires uids (NOT accessions).
ensembl_get_sequence with gene ID (ENSG) + type != "genomic" requires multiple_sequences=true. Use transcript IDs (ENST) for specific sequences.

Recipe: Get mRNA for a human gene

NCBI_search_nucleotide(organism="Homo sapiens", gene="BRCA1", seq_type="mRNA", limit=5)
NCBI_fetch_accessions(uids=[first_uid]) -> accession
NCBI_get_sequence(accession="NM_007294", format="fasta")

Phase 3: Protein Sequence Retrieval

UniProt_get_sequence_by_accession: accession (string REQUIRED, e.g., "P04637"). Returns {result: "MEEPQSDP..."}. Note: response key is result, NOT data.

EnsemblSeq_get_id_sequence: ensembl_id (string REQUIRED, e.g., "ENSP00000269305"), type ("protein"/"cdna"/"cds"). Returns {status, data: {ensembl_id, molecule, sequence, sequence_length}}.

UniProt_get_entry_by_accession: accession (string REQUIRED). Full protein annotation.

Gotchas:

UniProt_get_sequence_by_accession returns {result: "..."}, not {status, data}.
For Ensembl protein seqs, use ENSP IDs. For cDNA/CDS, use ENST IDs.
To find UniProt accession from gene: use NCBIDatasets_get_gene_by_symbol (has cross-refs).

Phase 4: Ortholog and Comparative Analysis

NCBIDatasets_get_orthologs: gene_id (string REQUIRED, NCBI Gene ID e.g., "7157"), page_size (int, default 20, max 100). Returns {status, data: [{gene_id, symbol, description, taxname, common_name, chromosomes}]}.

NCBIProtein_get_summary: id (string REQUIRED, GI number or accession). Returns protein title, organism, length.

Gotcha: NCBIDatasets_get_orthologs requires NCBI Gene ID (numeric string), not gene symbol or Ensembl ID. Resolve via Phase 1 first.

Recipe: Compare orthologs

NCBIGene_search(term="TP53[Symbol] AND Homo sapiens[Organism]") -> "7157"
NCBIDatasets_get_orthologs(gene_id="7157", page_size=10) -> mouse Trp53, rat Tp53, etc.

Phase 5: Domain Architecture and Homology

InterPro_get_entries_for_protein: accession (UniProt ID). Returns InterPro domain/family/superfamily entries with positions.

Pfam_get_protein_annotations: accession (UniProt ID). Returns Pfam domain hits with exact residue coordinates and E-values. For a raw sequence with no UniProt accession yet (e.g. a novel or designed sequence), EBI_scan_pfam_domains runs the same Pfam scan directly on the sequence instead.

BLAST_protein_search: sequence (amino acid string), database (default "swissprot"), limit. Returns homologs with alignment scores, identity, E-values. If BLAST finds nothing (remote homology, low sequence identity), EBI_profile_search (method='phmmer' default, or 'psiblast') uses profile/HMM methods that are more sensitive for distant relationships BLAST's pairwise scoring misses.

EBI_pairwise_align: exactly two sequences (algorithm='needle' for global, 'water' for local), returns percent identity/similarity/gaps and the alignment itself -- no existing tool in this skill does direct pairwise alignment.

EBI_translate_sequence: nucleotide-to-protein, protein-to-nucleotide (back-translation), or 6-frame translation with ORF detection. DNA_translate_reading_frames above only does forward 3-frame (no reverse strand, no back-translation) -- use EBI's tool specifically for reverse-strand frames or when going protein-to-DNA.

EBI_predict_membrane_topology: predicts transmembrane helices and signal peptides directly from a protein sequence (Phobius). Distinct from proteins_api_get_features's TRANSMEM annotations (see "Reasoning for Protein Feature Questions" below) -- that's looking up already-annotated topology for a known UniProt entry; this predicts topology for a sequence that may have none.

EnsemblCompara_get_orthologues: gene (gene symbol, e.g., "CFTR"), species (e.g., "human"). User-friendly alternative to NCBIDatasets_get_orthologs — accepts gene symbols directly.

Phase 6: Variant and Clinical Context

EnsemblVEP_annotate_hgvs: hgvs_notation (e.g., "NM_000492.4:c.1521_1523del"). Returns consequence, protein impact, genomic coordinates.

ClinVar_search_variants: gene (gene symbol). Returns variant count and IDs for clinical significance lookup.

PubMed_search_articles: query, limit. Literature context for gene/variant findings.

Tool Parameter Quick Reference

Tool	Correct Param	Common Mistake
NCBIGene_search	`term` (with [Symbol] syntax)	`query` or `gene`
NCBIGene_get_summary	`id` (string)	Integer type
NCBI_fetch_accessions	`uids` (array)	`accessions`
NCBI_get_sequence	`accession` (string)	Passing UID
NCBIDatasets_get_orthologs	`gene_id` (string)	Gene symbol
EnsemblSeq_get_id_sequence	`ensembl_id`	`id`
ensembl_get_sequence	`id` + `multiple_sequences`	Omitting multiple_sequences for gene+CDS
UniProt_get_sequence_by_accession	`accession`	Response is `result` not `data`

Fallbacks

Gene not found -> try NCBIDatasets_get_gene_by_symbol with explicit taxon
No accessions from search -> broaden query (remove strain/seq_type filters)
Ensembl error for gene+CDS -> use transcript ID (ENST) or set multiple_sequences=true
UniProt accession unknown -> NCBIDatasets_get_gene or UniProt_search for cross-refs
Ortholog search empty -> verify gene_id is numeric NCBI Gene ID

Sequence Analysis Reasoning (CRITICAL)

LOOK UP DON'T GUESS -- always fetch sequences, coordinates, and domain boundaries from databases. Do not reconstruct them from memory.

When to Use Which Tool

Question Type	Tool Choice	Why
"Find similar sequences"	BLAST_protein_search	Homology search against databases; returns E-values and identity
"What domains does this protein have?"	InterPro_get_entries_for_protein or Pfam_get_protein_annotations	Domain architecture with exact residue coordinates
"Get the sequence of gene X"	NCBI_search_nucleotide -> NCBI_get_sequence	Nucleotide retrieval by gene name
"Compare orthologs"	NCBIDatasets_get_orthologs or EnsemblCompara_get_orthologues	Cross-species gene comparison
"What is the protein impact of variant X?"	EnsemblVEP_annotate_hgvs	Consequence prediction with protein coordinates
"Align two sequences"	BLAST (pairwise)	Quick pairwise comparison with scoring

Reading Frame Selection Strategy

When translating a DNA sequence to protein:

Do NOT guess the reading frame -- preferred: use DNA_translate_reading_frames tool; fallback: translate_dna.py which tries all 3 frames automatically
The correct frame is the one with the LONGEST open reading frame (no premature stops)
If the sequence starts with ATG, frame 1 is likely correct -- but verify
If all 3 frames have early stop codons, the sequence may be: (a) non-coding, (b) reversed, or (c) contains sequencing errors. Try reverse complement first.

Protein Domain Interpretation

When asked about protein function or structure:

Get domain architecture first: InterPro_get_entries_for_protein returns all annotated domains with positions
Domain families indicate function: Kinase domain = phosphorylation activity; SH2 domain = phosphotyrosine binding; zinc finger = DNA binding
Variants in conserved domains are more likely pathogenic than those in linker regions
LOOK UP domain boundaries from the database -- do not estimate positions from memory

Reasoning for Protein Feature Questions

When asked "how many X residues in region Y of protein Z":

Identify the correct protein — Gene names are ambiguous. GABAA has many subunits (GABRA1, GABRB2, GABRR1...). Read the question carefully for the specific subunit. Use proteins_api_search with gene name + "human" to find the right accession.
Find the region boundaries — Use proteins_api_get_features with the accession to get annotated domains (TRANSMEM, DOMAIN, REGION). Don't guess positions — get them from the database.
Count residues in the region — Fetch the sequence, extract the region, count. WRITE Python code for this — don't try to count manually.
- Residue Counting Strategy: python3 skills/tooluniverse-sequence-analysis/scripts/sequence_tools.py --type count_region --accession P24046 --start 318 --end 440 --residue C
- For residue counting questions, ALWAYS use the script or sequence[start:end].count('C'). Do NOT estimate or count from memory.
Account for multimers — READ THE QUESTION for "homomeric", "pentamer", "tetramer", "dimer". If the question asks about a homomeric receptor (e.g., "homomeric GABAAρ1"), every subunit is identical. Count the residues in ONE subunit, then multiply:
- Homomeric pentamer (most ligand-gated ion channels like GABAA ρ1): × 5
- Homotetramer (many ion channels): × 4
- Homodimer: × 2 If the question says "in the TM3-TM4 linker domains" (plural), it means across all subunits in the complex.

Bundled Computation Scripts

Never manually count residues, compute GC%, or write reverse-complement logic inline. Run these scripts instead — they are tested and handle edge cases.

biology_facts.py — Biology reference lookup

Script: skills/tooluniverse-sequence-analysis/scripts/biology_facts.py

Use this script to look up commonly-confused biology facts instead of relying on memory. It covers receptor types, ion channel stoichiometry, neurotransmitters, immune cell markers, and gene naming confusions.

python3 skills/tooluniverse-sequence-analysis/scripts/biology_facts.py --type receptor --name "GABAA"
python3 skills/tooluniverse-sequence-analysis/scripts/biology_facts.py --type ion_channel --name "NMDA"
python3 skills/tooluniverse-sequence-analysis/scripts/biology_facts.py --type gene_confusion --name "GABRA1"
python3 skills/tooluniverse-sequence-analysis/scripts/biology_facts.py --type receptor  # list all entries

Types: receptor (stoichiometry, pharmacology), ion_channel (subunit arrangement), neurotransmitter (synthesis, receptors), immune_cell (markers, lineage), gene_confusion (commonly mixed-up genes like GABRA1 vs GABRR1).

Mandatory use: any question about receptor type/stoichiometry, immune cell markers, or gene name disambiguation.

amino_acids.py — Codon table, amino acid properties, wobble pairing

Script: skills/tooluniverse-sequence-analysis/scripts/amino_acids.py

Use this script for any question about the genetic code, codon degeneracy, amino acid chemistry, codon usage bias, or tRNA wobble pairing. All outputs are JSON. Its codon-usage frequencies are human-only ("human codon usage %") -- for any other organism (e.g. codon-optimizing for E. coli, yeast, or CHO expression), use CodonUsage_get_table/CodonUsage_get_optimal_codons(taxid=...) instead, which cover any NCBI-taxid organism.

python3 skills/tooluniverse-sequence-analysis/scripts/amino_acids.py --type codon_table
python3 skills/tooluniverse-sequence-analysis/scripts/amino_acids.py --type amino_acid --name "Cysteine"
python3 skills/tooluniverse-sequence-analysis/scripts/amino_acids.py --type amino_acid --code C
python3 skills/tooluniverse-sequence-analysis/scripts/amino_acids.py --type amino_acid --code TRP
python3 skills/tooluniverse-sequence-analysis/scripts/amino_acids.py --type amino_acid            # list all 20
python3 skills/tooluniverse-sequence-analysis/scripts/amino_acids.py --type count_codons --sequence "ATGCCCAAATTT..."
python3 skills/tooluniverse-sequence-analysis/scripts/amino_acids.py --type wobble --anticodon "GAU"
python3 skills/tooluniverse-sequence-analysis/scripts/amino_acids.py --type wobble --anticodon "IAU"

Modes:

`--type`	What it returns	Key fields
`codon_table`	All 64 codons grouped by amino acid	degeneracy, codons, human codon usage %, stop codon names, degeneracy distribution (1/2/3/4/6)
`amino_acid`	Properties of one or all amino acids	name, one_letter, three_letter, mw_da, pKa_side_chain, polarity, charge_ph7, hydrophobicity_index (Kyte-Doolittle), backbone_pKa, codons, degeneracy, rare_codons_le15pct
`count_codons`	Codon frequency analysis for a DNA sequence	codon_counts with AA annotation and human usage freq, amino_acid_composition, rare_codons_present
`wobble`	Codons recognised by a given anticodon	recognised_codons (RNA+DNA form, AA), synonymous_only, wobble rule explanation

When to use (mandatory):

Any question about how many codons encode a given amino acid (degeneracy)
Any question about rare vs. common codons for protein expression optimisation
Any question about tRNA anticodon recognition / wobble base pairing
Any question about amino acid physical-chemical properties (MW, pKa, hydrophobicity, polarity, charge)
Any question about the names of stop codons (Amber/Ochre/Opal)
Before manually stating codon degeneracy — verify with codon_table

Wobble rules: I pairs U/C/A (3 codons); G pairs U/C; U pairs A/G; C pairs G only; A pairs U only (rare). Use --type wobble --anticodon "GAU" to verify.

Amino acid lookup: accepts full name (--name "Cysteine"), 1-letter (--code C), or 3-letter (--code CYS).

Codon-Anticodon Matching Reasoning (CRITICAL for tRNA problems)

When solving "which codons does this tRNA recognize" or "which tRNA reads this codon":

Anticodon is written 3'->5' but conventionally listed 5'->3'. The FIRST position of the anticodon (5' end) is the WOBBLE position and pairs with the THIRD position of the codon (3' end).
Anticodon-codon pairing is ANTIPARALLEL: anticodon 5'-X-Y-Z-3' pairs with codon 3'-X'-Y'-Z'-5' (i.e., codon 5'-Z'-Y'-X'-3').
Wobble position rules (anticodon 5' base -> codon 3' base it can pair with):
- C -> G only (1 codon)
- A -> U only (1 codon; rare in bacteria, common in mitochondria)
- U -> A or G (2 codons)
- G -> C or U (2 codons)
- I (inosine, deaminated A) -> U, C, or A (3 codons)
Minimum tRNA set: Because I reads 3 bases and G/U each read 2, a 4-codon family (e.g., GCN = Ala) needs only 2 tRNAs: one with I at wobble position (reads 3 of 4 codons) and one with C or U at wobble (reads the remaining 1-2).
ALWAYS use the script: python3 skills/tooluniverse-sequence-analysis/scripts/amino_acids.py --type wobble --anticodon "IAU" to verify rather than reasoning from memory.

translate_dna.py — DNA to protein translation

Preferred: use DNA_translate_reading_frames tool (via MCP/SDK) with sequence parameter. Fallback: run translate_dna.py directly.

python3 skills/tooluniverse-sequence-analysis/scripts/translate_dna.py "ATGCCC..."

Tries all 3 reading frames, picks longest ORF automatically.

sequence_tools.py — Residue counting, GC content, reverse complement, stats

Script: skills/tooluniverse-sequence-analysis/scripts/sequence_tools.py

Preferred: Use ToolUniverse tools (via MCP/SDK) instead of the script:

Sequence_count_residues tool -- Count residues in a sequence or region. Fallback: sequence_tools.py --type count_residues or --type count_region
Sequence_gc_content tool -- GC% of DNA. Fallback: sequence_tools.py --type gc_content
Sequence_reverse_complement tool -- DNA reverse complement. Fallback: sequence_tools.py --type reverse_complement
Sequence_stats tool -- Auto-detect type, length, MW. Fallback: sequence_tools.py --type stats

Fallback script modes (use --type):

count_residues: Count residue in full sequence. --sequence "ACDE..." --residue C
count_region: Count in region (1-based inclusive). --sequence "MAC..." --start 5 --end 20 --residue C OR --accession P24046 --start 318 --end 440 --residue C (fetches from UniProt live)
gc_content: GC% of DNA. --sequence "ATGCGATCG"
reverse_complement: DNA reverse complement. --sequence "ATGCGATCG"
stats: Auto-detect DNA/RNA/Protein, compute length, MW for protein. --sequence "ATGCG..."

ALWAYS use count_region --accession when the user gives a UniProt accession + region -- do not count manually.

Interpretation Framework

Sequence Quality Assessment

Indicator	High Quality	Acceptable	Caution
RefSeq status	NM_/NP_ (curated)	XM_/XP_ (predicted)	No RefSeq (GenBank only)
Sequence version	Latest version (.N)	Previous version	Removed/replaced
Annotation	Reviewed (UniProt Swiss-Prot)	Unreviewed (TrEMBL)	No annotation
Gene symbol	HGNC approved	Alias/synonym	Locus tag only

Synthesis Questions

Is this the correct sequence? (verify organism, gene symbol, isoform)
Is it the canonical isoform? (RefSeq MANE Select or UniProt canonical)
How well-annotated is it? (SwissProt > TrEMBL > GenBank predicted)
Are there known variants? (ClinVar pathogenic variants in this sequence)

Answer Formatting (CRITICAL)

TRIM YOUR ANSWER: If the question asks "what protein", answer with JUST the protein name. Do not add parenthetical abbreviations, descriptions, or qualifications. Example: answer "Glucose-6-phosphate 1-dehydrogenase", NOT "Glucose-6-phosphate 1-dehydrogenase (G6PD, EC 1.1.1.49)". When identifying a protein from a sequence, use BLAST/UniProt and report the top hit name exactly as it appears in the database — no embellishment.

Peptide & Foldamer Structure

Alpha-peptide helices: alpha-helix (3.6 res/turn, i->i+4 H-bonds), 3_10-helix (3 res/turn, i->i+3), pi-helix (4.4 res/turn, i->i+5).
Beta-peptide helices: named by H-bond ring size. 14-helix (i->i+2, 14-membered rings), 12-helix, 10-helix, 8-helix.
Beta-amino acid ring size determines helix type: 4-membered cyclic constraint -> 10-helix; 5-membered (e.g., ACPC) -> 12-helix; 6-membered (e.g., ACHC) -> 14-helix. Acyclic beta3-residues default to 14-helix.
Mixed alpha/beta foldamers (1:1 alternation): form 11-helix (i->i+3, 11-atom rings) or 14/15-helix (i->i+4, alternating 14- and 15-atom rings). Longer sequences prefer the 14/15-helix.
Key rule: the number in the helix name = number of atoms in the hydrogen-bonded ring.
Cyclic beta-amino acids (ACPC, ACHC) constrain backbone torsion angles, favoring specific helix types over acyclic residues.

Limitations

ensembl_get_sequence gene IDs + non-genomic type need multiple_sequences=true
NCBIDatasets_get_orthologs requires NCBI Gene ID (not symbol); UniProt returns canonical isoform only

When it activates

Only when you call it by name. Your agent never loads it on its own. Use for sequence retrieval, sequence comparison, FASTQ QC analysis, and read alignment pre-processing.

About this skill

On-demandOn-demand: invoked explicitly — never occupies your agent's context until called.

Research pack

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v5 — 9/22/2026

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

mims-harvard / ToolUniverse

skills/tooluniverse-sequence-analysis/SKILL.md

View on GitHub ↗

License Apache-2.0Imported 9/22/2026Version v2Bundled 4 files · scripts

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

Other skills for the same job

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

mims-harvard/tooluniverse-sequence-analysisBiological sequence analysis — gene/protein sequence retrieval (NCBI, Ensembl, UniProt), nucleotide/protein search, ortholog discovery, and FASTQ QC + alignment workflows (Trimmomatic, BWA, samtools, coverage depth). Use for sequence retrieval, sequence comparison, FASTQ QC analysis, and read alignment pre-processing.Near-identical description-33 pts pass-rate lift

Related skills

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

From the same source and nearby in the catalogue.

Ready to supercharge your AI agents?

Safety

What it changes

Track record

Use this skill

Biological Sequence Analysis

⚠️ TOP-OF-MIND RULE: Trimmomatic "reads completely discarded" = F + R + 2*D, summed across samples

RULE ZERO — Check for pre-computed results FIRST

FASTQ QC, Trimmomatic, and read alignment (when raw reads are present)

Trimmomatic PE — counting "completely discarded" reads

Coverage depth (samtools depth / mosdepth)

When to Use

Workflow

Phase 1: Gene Identification and Summary

Phase 2: Nucleotide Sequence Search and Retrieval

Recipe: Get mRNA for a human gene

Phase 3: Protein Sequence Retrieval

Phase 4: Ortholog and Comparative Analysis

Recipe: Compare orthologs

Phase 5: Domain Architecture and Homology

Phase 6: Variant and Clinical Context

Tool Parameter Quick Reference

Fallbacks

Sequence Analysis Reasoning (CRITICAL)

When to Use Which Tool

Reading Frame Selection Strategy

Protein Domain Interpretation

Reasoning for Protein Feature Questions

Bundled Computation Scripts

biology_facts.py — Biology reference lookup

amino_acids.py — Codon table, amino acid properties, wobble pairing

Codon-Anticodon Matching Reasoning (CRITICAL for tRNA problems)

translate_dna.py — DNA to protein translation

sequence_tools.py — Residue counting, GC content, reverse complement, stats

Interpretation Framework

Sequence Quality Assessment

Synthesis Questions

Answer Formatting (CRITICAL)

Peptide & Foldamer Structure

Limitations

When it activates

SkillSafety

Source

Other skills for the same job

Related skills

More in the registry

⚠️ TOP-OF-MIND RULE: Trimmomatic "reads completely discarded" = `F + R + 2*D`, summed across samples