Ready to supercharge your AI agents?

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

Create Adapter · marcus | DecimalAI

Agent Toolingby marcus

Create Adapter

Create conversation adapters for importing AI chat history from different tools (Claude Code, Cursor, Warp, Codex, etc.). Covers the adapter.Adapter interface, caching strategies, incremental parsing, watch/FD management, and performance standards.

Never made a case worseMaintained

SkillScore

SkillScore100/100

Test results 100

Safety

3/3 safety checks passed

Code scan
Intent
Content

What it changes

45%100%of test cases passed +55 pts

Without skill10 of 22

With skill22 of 22

12 fixed10 passed either way0 still failing0 broke

Measured on gemini-3.6-flash8/16/2026v221 comparable

Track record

Agent runs: 0; All time

See one task, with and without the skill

One test taskWhen sending file change events over an adapter watch channel in Sidecar, how should channel sends be written to avoid blocking the watcher goroutine when consumer channels are full?

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 marcus-create-adapter
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("marcus-create-adapter")
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 marcus-create-adapter

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/marcus-create-adapter/SKILL.md https://app.decimal.ai/s/marcus-create-adapter/SKILL.md

Copy SKILL.md

Save it in your project, as

.claude/skills/marcus-create-adapter/SKILL.md

This gets SKILL.md only. The skill's 2 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 marcus-create-adapter

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("marcus-create-adapter")
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/marcus-create-adapter/SKILL.md

Give it to your agent

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

skill = Path(".agents/skills/marcus-create-adapter/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 2 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 marcus-create-adapter

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("marcus-create-adapter")
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/marcus-create-adapter/SKILL.md

Give it to your agent

from pathlib import Path
from agents import Agent, Runner

skill = Path(".agents/skills/marcus-create-adapter/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 2 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 marcus-create-adapter

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("marcus-create-adapter")
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/marcus-create-adapter/SKILL.md

Give it to your agent

from pathlib import Path
from openai import OpenAI

client = OpenAI()
skill = Path(".agents/skills/marcus-create-adapter/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 2 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 marcus-create-adapter

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("marcus-create-adapter")
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/marcus-create-adapter/SKILL.md

Give it to your agent

from pathlib import Path
import anthropic

client = anthropic.Anthropic()
skill = Path(".agents/skills/marcus-create-adapter/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 2 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 marcus-create-adapter

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("marcus-create-adapter")
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/marcus-create-adapter/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/marcus-create-adapter/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 2 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 marcus-create-adapter

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("marcus-create-adapter")
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/marcus-create-adapter/SKILL.md

Give it to your agent

from pathlib import Path
from langchain.agents import create_agent

skill = Path(".agents/skills/marcus-create-adapter/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 2 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 marcus-create-adapter

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("marcus-create-adapter")
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/marcus-create-adapter/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 2 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 marcus-create-adapter

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("marcus-create-adapter")
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/marcus-create-adapter/SKILL.md

Give it to your agent

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

skill = Path(".agents/skills/marcus-create-adapter/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 2 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 marcus-create-adapter

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("marcus-create-adapter")
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/marcus-create-adapter/SKILL.md

Give it to your agent

from pathlib import Path
from pydantic_ai import Agent

skill = Path(".agents/skills/marcus-create-adapter/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 2 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 marcus-create-adapter

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("marcus-create-adapter")
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/marcus-create-adapter/SKILL.md

Give it to your agent

from pathlib import Path
from deepagents import create_deep_agent

skill = Path(".agents/skills/marcus-create-adapter/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 2 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 marcus-create-adapter

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("marcus-create-adapter")
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/marcus-create-adapter/SKILL.md

Give it to your agent

import asyncio
from pathlib import Path
from agent_framework import Agent

skill = Path(".agents/skills/marcus-create-adapter/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 2 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 marcus-create-adapter

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("marcus-create-adapter")
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/marcus-create-adapter/SKILL.md

Give it to your agent

from pathlib import Path
from strands import Agent

skill = Path(".agents/skills/marcus-create-adapter/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 2 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 marcus-create-adapter

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/marcus-create-adapter/SKILL.md https://app.decimal.ai/s/marcus-create-adapter/SKILL.md

Copy SKILL.md

Save it in your project, as

.agents/skills/marcus-create-adapter/SKILL.md

This gets SKILL.md only. The skill's 2 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/marcus-create-adapter/SKILL.md

Read the summary as JSON

curl -fsSL https://app.decimal.ai/s/marcus-create-adapter.json

This gets SKILL.md only. The skill's 2 other files are not included.

Other ways to add it

Other setups docs

You don't need a DecimalAI account.

Version: v2, tested 8/16/2026
Source: marcus/sidecar ↗
License: MIT

Manage it in DecimalAI

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

Use this skill

SKILL.md

Create Adapter

Why Performance Matters

Adapters are the largest performance risk in Sidecar. Conversations refresh on watch events in a hot path that runs continuously during active sessions:

watch event -> coalescer -> session refresh -> adapter.Sessions() -> metadata parsing

If an adapter does full directory scans and full-file reparses on every change, CPU and FD usage spike quickly.

Reference Adapters

Study these before writing a new adapter:

internal/adapter/claudecode - Incremental JSONL parsing, targeted refresh
internal/adapter/codex - Directory cache, two-pass metadata parsing, global watch scope
internal/adapter/cursor - SQLite/WAL-aware cache invalidation, FD-safe DB access
internal/adapter/pi - Global scope, JSONL, CWD-based filtering, session classification, message prefix stripping

Required Interface

All adapters implement adapter.Adapter:

type Adapter interface {
    ID() string
    Name() string
    Icon() string
    Detect(projectRoot string) (bool, error)
    Capabilities() CapabilitySet
    Sessions(projectRoot string) ([]Session, error)
    Messages(sessionID string) ([]Message, error)
    Usage(sessionID string) (*UsageStats, error)
    Watch(projectRoot string) (<-chan Event, io.Closer, error)
}

Required Session Fields

Every session from Sessions() must set:

ID, Name
AdapterID, AdapterName, AdapterIcon
CreatedAt, UpdatedAt
MessageCount, FileSize

FileSize is used for dynamic debounce and huge-session auto-reload protection.

Treat source identity separately from lineage. Use the source's durable thread/session ID for Session.ID; parent, root, fork, or lineage IDs describe relationships and must not collapse distinct sessions. Decode metadata fields defensively when the source has emitted multiple shapes over time (for example, a string in one version and an object in another).

Path and Watch Strategy

Set Session.Path only when Sidecar should use tiered file watching for that adapter:

File-based append-only (JSONL/log): set Path to absolute file path — this opts into TieredWatcher with HOT/COLD/FROZEN tiers
DB/WAL adapters (Cursor, Warp, Kiro): prefer adapter-specific Watch() with WAL-aware invalidation; do not set Path unless tiered watching covers your write surface

FROZEN tier: File-based sessions with Path set automatically benefit from the FROZEN tier. Sessions unchanged for 24 hours (FrozenThreshold) are excluded from cold polling entirely — zero syscalls. They unfreeze when promoted to HOT (e.g., user selects the session). This is critical for adapters with thousands of session files; without it, pollColdSessions() does one os.Stat() per file every 30 seconds.

Performance Standards

1) Cache metadata and messages aggressively

Minimum cache keys:

Metadata: path + size + modTime
Messages: path + size + modTime
SQLite/WAL: include WAL size+mtime in the key

Use bounded LRU behavior for every cache and index. Prune stale paths. Assume caches evict independently: a hit in one cache must restore any derived state required by another, or the authoritative source must remain available so eviction cannot change results such as aggregate usage or ID-to-path resolution.

2) Incremental parsing for append-only formats

For JSONL/event-log adapters:

Cache last parsed byte offset
Parse only appended bytes
Fall back to full parse on shrink/rotation/corruption
Preserve immutable head metadata from prior parse

3) Two-pass metadata for large files

When incremental metadata parse is impractical:

Head pass: ID, CWD, first user message, first timestamp
Tail pass: latest timestamp, token totals
Skip middle of large files

When the source owns a metadata index, prefer its read-only index over scanning large event logs. Probe the schema and required columns before use, open it read-only with bounded/FD-safe access, and fall back to event-log discovery when it is missing, locked, or incompatible. The source index is an adapter seam, not a second source of truth to mutate.

4) Avoid repeated expensive path work

Resolve project path once per Sessions() call (Abs/EvalSymlinks), reuse for all matches.

5) Return defensive copies from caches

Never return cache-owned slices/maps directly. Copy message/session structures to avoid mutation bugs.

6) Keep DB access FD-safe

For SQLite adapters:

Open read-only (mode=ro)
SetMaxOpenConns(1), SetMaxIdleConns(0)
Close rows and DB handles promptly
Avoid multiple DB connections per Messages() call

7) Preserve aggregate facts across incremental loads

Usage and similar cumulative facts may arrive as repeated totals or deltas. Define the source semantics, retain the authoritative aggregate across incremental parsing, and include all components the source exposes. Do not reconstruct a partial aggregate from whichever message cache entry survived eviction.

Watching and FD Management

1) Prefer directory-level watches

Do not watch per-session files when directory-level watch gives equivalent signals.

2) Implement watch scope

If adapter watches a global path (same location regardless of worktree):

func (a *Adapter) WatchScope() adapter.WatchScope {
    return adapter.WatchScopeGlobal
}

This prevents duplicate watchers across worktrees.

3) Always emit SessionID when known

Watch events should include session ID for targeted refresh (avoids full reloads).

4) Debounce and non-blocking sends

Debounce bursty write events
Use buffered channels
Non-blocking sends: select { case ch <- evt: default: }

5) Leverage FROZEN tier for file-based adapters

File-based adapters that set Session.Path get TieredWatcher's three-tier system (HOT → COLD → FROZEN). Sessions unchanged for 24h are frozen and cost zero polling overhead. This is the primary defense against CPU spikes with thousands of session files. If your adapter has file-based sessions, always set Path — the FROZEN tier scales automatically.

6) Ensure cleanup

All watcher paths must close cleanly on plugin stop. No goroutine or FD leaks.

7) Combine known-file and discovery watching when needed

Tiered watching of known Session.Path values handles cheap appends, but it cannot discover a project's first session or a new time-partitioned directory. Global file adapters may need both known-file watching and one adapter-native discovery watcher. Watch creation at every directory level that can appear later (including month/year rollover), and keep discovery project-filtered.

8) Resolve path IDs through the adapter

Do not assume a new file's basename is its session ID. If identity lives in metadata, implement SessionPathResolver so watch events carry the same durable ID returned by Sessions().

Adapter Call Lifecycle

Some global adapters maintain caches and indexes across calls. Consumers must serialize calls to a stateful adapter, make gate admission and work lifecycle/epoch cancellable, and reject stale results after project switches or shutdown. A slow Sessions() result must remain observable and eventually load (with a visible loading state); never time it out, silently discard it, and leave its goroutine running. Global targeted refresh must admit only sessions already belonging to the current project; unknown IDs require a project-filtered discovery pass.

Message and Content Rendering

Adapters must provide rich structured content for Conversation Flow UI.

Required message mapping

Map source records to:

Message.Role, Message.Content, Message.ContentBlocks
Message.ToolUses (legacy compatibility)
Message.ThinkingBlocks (if available)
Message.Model when available

Tool linking rule

Use consistent ToolUseID for tool_use and tool_result blocks. If incremental parsing is used, preserve pending tool-link state across cache updates.

Optional Interfaces

TargetedRefresher

type TargetedRefresher interface {
    SessionByID(sessionID string) (*Session, error)
}

Reduces refresh from O(N sessions) to O(1). Implement when adapter can resolve a session directly.

ProjectDiscoverer

Implement when source format allows discovery of sessions beyond current git worktrees.

SessionPathResolver

Implement when a file path alone does not encode the source's durable session ID. Tiered and discovery watchers use it to turn new or changed paths into targeted refresh events.

ProjectDiscoveryWatcher

Implement for global sources that must discover the first matching session even when Sessions() initially returns no known files. Share only one global watcher per adapter, and ensure its events trigger a project-filtered load before any session is admitted.

Error Handling

Detect(): return (false, nil) for missing data directories
Sessions(): skip corrupt/unreadable entries and continue; hard-fail only on systemic errors
Messages(): return nil, nil for missing session files; fail on parse errors
Watch(): return (nil, nil, err) when watch setup fails
Slow calls: surface loading/error state and accept the eventual result; do not translate a consumer timeout into a false empty history

Benchmark Targets

New adapters should meet these performance targets:

Messages() full parse (~1MB): under 50ms
Messages() incremental append: under 10ms
Messages() cache hit: under 1ms
Sessions() on 50 session files: under 50ms

Also benchmark realistic source shapes: hundreds or thousands of indexed sessions, a large live- scale transcript, cache hits, and incremental appends. Record fixture size and session/event count with the result so a tiny synthetic benchmark cannot hide discovery or parsing regressions.

Testing Requirements

Required tests for every new adapter:

Relative vs absolute project path behavior in Detect()/Sessions()
Sessions() sorted by UpdatedAt desc
Required session fields populated (Adapter*, FileSize, Path when applicable)
Cache hit behavior (no reparsing on unchanged files)
File growth behavior (incremental parse path)
File shrink/rotation behavior (fallback full parse)
Tool use/result linking (including incremental append cases)
Incremental ContentBlocks tool-use/result ID parity, not only legacy ToolUses
Aggregate usage across multiple events, cache-write/read components, and independent cache eviction
Watcher event emission includes SessionID
Watcher cleanup (no leaked closers)
Repeated-call FD stability
Zero-history project followed by its first session creation
Time-partition rollover (for example, a new month under an existing year)
Global discovery and targeted refresh remain isolated across two projects/worktrees

Run tests:

go test ./internal/adapter/<adapter> -run .
go test ./internal/adapter/<adapter> -bench . -benchmem

PR Compliance Checklist

A) Correctness

Full adapter.Adapter contract implemented
Sessions() sets required identity and timestamp fields
Durable source identity is distinct from parent/root lineage; historical metadata shapes parse
FileSize populated for every session
Path strategy explicit and correct for adapter type
Message role/content mapping correct
ContentBlocks include text/tool/thinking data
Tool result linking correct (ToolUseID parity)

B) Performance

Metadata cache implemented and bounded
Message cache implemented and bounded
Every auxiliary index/aggregate cache is bounded and correct under independent eviction
Incremental parse or two-pass strategy implemented
Source-owned metadata indexes are read-only, schema-probed, and have a safe fallback
No repeated Abs/EvalSymlinks in per-session loops
No duplicate parsing for single-pass data
Benchmarks added with realistic fixtures

C) FD / Watching

Directory-level watches preferred
Global adapters implement WatchScopeProvider
Global discovery is project-isolated, including zero-history first creation
Watch events include SessionID
Metadata-backed IDs resolve through SessionPathResolver
Debounce + buffered + non-blocking send pattern
DB adapters account for WAL in invalidation/watch
Known-file and discovery watches cover time-partition rollover without duplication
Watchers and goroutines close cleanly

D) Integration

Adapter registered via register.go and main import
Search uses adapter Messages() path
Large-session behavior validated (FileSize-driven)
Slow/global calls are serialized, lifecycle-cancellable, and never discarded as empty

Session Classification

Adapters can classify sessions by setting SessionCategory on adapter.Session. The conversations plugin supports category filtering (f menu: i/r/s keys) and a quick toggle (C key).

Category Constants

Defined in internal/adapter/adapter.go:

adapter.SessionCategoryInteractive — user-initiated interactive sessions
adapter.SessionCategoryCron — automated/scheduled sessions
adapter.SessionCategorySystem — system/gateway sessions

Implementation Guidelines

Classify during metadata parsing (zero extra I/O) — extract category from the first user message or session header
Only set SessionCategory if the adapter has meaningful categories. Don't set it if all sessions are the same type
If the category filter is active and SessionCategory is empty, sessions pass through (non-breaking for adapters that don't classify)
Gateway/system messages may need special classification — e.g., "System: WhatsApp gateway connected" is actually interactive, not system. Check for known preamble patterns before defaulting to system category

Example (from Pi adapter)

func extractSessionMetadata(firstUserMessage string) (category, cronJobName, sourceChannel string) {
    if strings.HasPrefix(firstUserMessage, "[cron:") {
        return adapter.SessionCategoryCron, extractCronJobName(firstUserMessage), ""
    }
    if strings.HasPrefix(firstUserMessage, "System:") {
        if strings.Contains(firstUserMessage, "WhatsApp gateway") {
            return adapter.SessionCategoryInteractive, "", "whatsapp"
        }
        return adapter.SessionCategorySystem, "", ""
    }
    return adapter.SessionCategoryInteractive, "", detectSourceChannel(firstUserMessage)
}

Rich Metadata Fields

Optional fields on adapter.Session for richer display and filtering:

CronJobName string — for cron/scheduled sessions; used as session name when set
SourceChannel string — for multi-channel adapters (e.g., "telegram", "whatsapp", "direct")

Optional field on adapter.Message:

SourceLabel string — per-message source attribution badge (e.g., "[TG] Marcus", "[WA]", "[cron] job-name")

Set these during parsing when the source format contains channel/origin metadata. The conversations plugin and conversation flow UI use these for display.

Message Content Cleaning

For adapters whose source format embeds structured prefixes in user messages (e.g., channel tags, cron headers), strip them during parsing to keep the conversation view clean.

Pattern

Extract metadata (source label, channel, category) from the raw message prefix
Strip the prefix from Message.Content and text ContentBlocks
Store the extracted label in Message.SourceLabel for badge display

// In processMessageLine for user messages:
content, _, _, contentBlocks := parseContent(raw.Message.Content)
sourceLabel := extractSourceLabel(content)    // "[TG] Marcus"
content = stripMessagePrefix(content)          // clean body only
for i := range contentBlocks {
    if contentBlocks[i].Type == "text" {
        contentBlocks[i].Text = stripMessagePrefix(contentBlocks[i].Text)
    }
}
msg := adapter.Message{
    Content:       content,
    ContentBlocks: contentBlocks,
    SourceLabel:   sourceLabel,
}

This keeps Content human-readable while preserving origin metadata in SourceLabel.

Global Adapter Gotchas

Lessons learned from building global-scope adapters (Pi, Codex):

CWD-based Project Filtering

Global adapters (WatchScopeGlobal) store sessions in a single directory regardless of project. They must filter by CWD matching projectRoot in Sessions():

Resolve projectRoot once per Sessions() call (Abs + EvalSymlinks)
Read CWD from the cheapest authoritative source (a read-only metadata index or event-log header), and bound any cache used to avoid full-file parses for non-matching sessions
Match with filepath.Rel — a session matches if its CWD is equal to or a subdirectory of the project root

Category Filter Interaction

The conversations plugin category filter only filters sessions that HAVE a SessionCategory set — empty passes through
Don't enable category filter by default in the plugin — it breaks non-classifying adapters
When adding classification to a new adapter, test that existing adapters without categories still display correctly

Project Switching

Global adapters need to handle project switching gracefully:

Each Sessions(projectRoot) result is project-filtered, even when adapter caches retain source-global metadata
Cross-project/path identity indexes may accumulate across serialized calls, but must be bounded; eviction must fall back to authoritative index/log lookup rather than changing resolution behavior
Project reinitialization cancels and replaces the prior watcher lifecycle/epoch; stale loads and events must be rejected
Cache entries keyed only by source path may be shared across projects, while membership in a returned session list remains project-specific

Watcher Lifecycle and Refresh

Create only one adapter-native global discovery watcher within the active lifecycle, alongside tiered watches for known files when applicable
Target refresh only for IDs already admitted to the current project's session list, and only through their owning adapter
Treat an unknown ID from a global watcher as discovery: run a project-filtered full load before admitting it
Ensure watch goroutines and queued calls use the current lifecycle context and cannot retain stale project state

Schema References

See references/cursor-db-format.md for Cursor's per-session SQLite database structure (Merkle tree blobs, hex-encoded metadata, WAL considerations).

See references/warp-sqlite-schema.md for Warp's single SQLite database structure (ai_queries, agent_conversations, blocks tables, protobuf tasks).

When it activates

Automatically, when a task matches. Use when creating a new adapter, modifying adapter behavior, or debugging adapter performance issues. See references/ for Cursor DB and Warp SQLite schema details.

About this skill

Automatic

Coding pack

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/10/2026

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

marcus / sidecar

.claude/skills/create-adapter/SKILL.md

View on GitHub ↗

License MITImported 8/10/2026Version v2Bundled 2 files

Illustrative Want to see what marcus/create-adapter 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

Create Adapter

Why Performance Matters

Reference Adapters

Required Interface

Required Session Fields

Path and Watch Strategy

Performance Standards

1) Cache metadata and messages aggressively

2) Incremental parsing for append-only formats

3) Two-pass metadata for large files

4) Avoid repeated expensive path work

5) Return defensive copies from caches

6) Keep DB access FD-safe

7) Preserve aggregate facts across incremental loads

Watching and FD Management

1) Prefer directory-level watches

2) Implement watch scope

3) Always emit SessionID when known

4) Debounce and non-blocking sends

5) Leverage FROZEN tier for file-based adapters

6) Ensure cleanup

7) Combine known-file and discovery watching when needed

8) Resolve path IDs through the adapter

Adapter Call Lifecycle

Message and Content Rendering

Required message mapping

Tool linking rule

Optional Interfaces

TargetedRefresher

ProjectDiscoverer

SessionPathResolver

ProjectDiscoveryWatcher

Error Handling

Benchmark Targets

Testing Requirements

PR Compliance Checklist

A) Correctness

B) Performance

C) FD / Watching

D) Integration

Session Classification

Category Constants

Implementation Guidelines

Example (from Pi adapter)

Rich Metadata Fields

Message Content Cleaning

Pattern

Global Adapter Gotchas

CWD-based Project Filtering

Category Filter Interaction

Project Switching

Watcher Lifecycle and Refresh

Schema References

When it activates

SkillSafety

Source

Related skills

More in the registry