Practical notes: Harness Engineering: The Naked Agent: Why Your Framework Hands

Use this as an operator-facing rebuild of the ideas in “Harness Engineering: The Naked Agent: Why Your Framework Hands You a Loop, Not a Harness — I”: clear stages, ordered code slots, and recovery notes that survive a handoff.

Part 1: A bare agent loop looks powerful until real traffic hits it. Here is why production failures usually come from the missing harness around the model, not the model itself.

The Part 1 A bare stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Document the happy path and the recovery path together. Retries, human gates, and dead-letter handling are part of the product, not later polish. Budget tokens per turn and per session. Agentic tools expand context aggressively; hard caps keep demos from becoming surprise invoices.

Most agent failures are not the model’s fault. They live in the missing discipline layer around it. Here is what an AI agent with no harness looks like in the Claude Agent SDK and LangChain Deep Agents, and the three specific ways it breaks under real traffic.

The Most agent failures are stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Prefer small, testable units over sprawling scripts. When a step fails, the failure should point at a single responsibility rather than a tangled pipeline. Budget tokens per turn and per session. Agentic tools expand context aggressively; hard caps keep demos from becoming surprise invoices.

The model is not the variable

The The model is not stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Treat this stage as a contract between inputs and validated outputs. Name the artifacts, define success checks, and refuse silent partial completion. Budget tokens per turn and per session. Agentic tools expand context aggressively; hard caps keep demos from becoming surprise invoices.

What “naked” actually means

The What naked actually means stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Record timings and token or query cost next to functional results. Cost visibility early prevents surprise bills when the path moves from demo to shared environments. Expose tools with narrow schemas and explicit side-effect labels. Hosts need to know which calls mutate state before they auto-approve.

import anthropic

client = anthropic.Anthropic()  # reads ANTHROPIC_API_KEY
TOOLS = [
    {"name": "search_flights",
     "description": "Search flights between two cities for a date.",
     "input_schema": {"type": "object", "properties": {
         "origin": {"type": "string"}, "destination": {"type": "string"},
         "date": {"type": "string", "description": "YYYY-MM-DD"}},
         "required": ["origin", "destination", "date"]}},
    {"name": "book_flight",
     "description": "Book a specific flight.",
     "input_schema": {"type": "object", "properties": {
         "flight_id": {"type": "string"}, "passenger_name": {"type": "string"}},
         "required": ["flight_id", "passenger_name"]}},
]
def run_naked(user_msg: str) -> str:
    messages = [{"role": "user", "content": user_msg}]
    while True:                                   # ① no iteration cap
        resp = client.messages.create(
            model="claude-sonnet-4-6", max_tokens=1024,
            tools=TOOLS, messages=messages,
        )
        if resp.stop_reason != "tool_use":
            return resp.content[0].text
        call = next(b for b in resp.content if b.type == "tool_use")
        result = dispatch(call.name, call.input)
                                             # ② direct side effect, no check
        messages.extend([
                                             # ③ whole history, every turn
            {"role": "assistant", "content": resp.content},
            {"role": "user", "content": [{"type": "tool_result",
                "tool_use_id": call.id, "content": result}]},
        ])

The naked agent in the Claude Agent SDK

The The naked agent in stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Keep configuration outside application code. Environment files, secret stores, and feature flags belong in one place operators can audit without reading the whole graph. Expose tools with narrow schemas and explicit side-effect labels. Hosts need to know which calls mutate state before they auto-approve. The The naked agent in stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Prefer small, testable units over sprawling scripts. When a step fails, the failure should point at a single responsibility rather than a tangled pipeline.

import asyncio
from claude_agent_sdk import (
    query, ClaudeAgentOptions, tool,
    create_sdk_mcp_server, AssistantMessage, ResultMessage,
)

@tool("search_flights", "Search flights between two cities for a date.",
      {"origin": str, "destination": str, "date": str})
async def search_flights(args):
                                              # ① no check that date exists
    hits = flights_api.search(**args)
    return {"content": [{"type": "text", "text": str(hits)}]}
@tool("book_flight", "Book a specific flight.",
      {"flight_id": str, "passenger_name": str})
async def book_flight(args):
                                               # ② destructive, ungated
    confirmation = flights_api.book(**args)
    return {"content": [{"type": "text", "text": confirmation}]}
server = create_sdk_mcp_server("travel", tools=[search_flights, book_flight])
async def main():
    options = ClaudeAgentOptions(
        mcp_servers={"travel": server},
        allowed_tools=["mcp__travel__search_flights",
                       "mcp__travel__book_flight"],
    )
    async for msg in query(prompt="Rebook this customer for March 32nd.",
                           options=options):
        if isinstance(msg, AssistantMessage):
            for b in msg.content:
                if hasattr(b, "text"):
                    print(b.text)
        elif isinstance(msg, ResultMessage):
            print("done:", msg.subtype)
                                              # ③ no state survives this run

The naked agent in LangChain Deep Agents

For the The naked agent in stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Treat this stage as a contract between inputs and validated outputs. Name the artifacts, define success checks, and refuse silent partial completion. Authenticate at the gateway and re-authorize at the data plane. A bearer token alone is not a tenancy boundary.

from langchain.tools import tool
from deepagents import create_deep_agent

@tool
def search_flights(origin: str, destination: str, date: str) -> str:
    """Search flights between two cities for a date (YYYY-MM-DD)."""
    return str(flights_api.search(origin, destination, date))
                                                # ① no date check
@tool
def book_flight(flight_id: str, passenger_name: str) -> str:
    """Book a specific flight."""
    return flights_api.book(flight_id, passenger_name)
                                                 # ② ungated side effect
agent = create_deep_agent(
                                                 # ③ the loop, no controls
    model="anthropic:claude-sonnet-4-6",
    tools=[search_flights, book_flight],
)
result = agent.invoke({"messages": [{"role": "user",
    "content": "Rebook this customer for March 32nd."}]})
print(result["messages"][-1].content)
# Ask a follow-up in a second invoke, and it starts from zero: no thread,
# no memory.

Watch it break three ways

For the Watch it break three stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Record timings and token or query cost next to functional results. Cost visibility early prevents surprise bills when the path moves from demo to shared environments. Authenticate at the gateway and re-authorize at the data plane. A bearer token alone is not a tenancy boundary.

Failure 1: a malformed argument reaches a destructive call

For the Failure 1 a malformed stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Keep configuration outside application code. Environment files, secret stores, and feature flags belong in one place operators can audit without reading the whole graph. Authenticate at the gateway and re-authorize at the data plane. A bearer token alone is not a tenancy boundary. For the Failure 1 a malformed stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Prefer small, testable units over sprawling scripts. When a step fails, the failure should point at a single responsibility rather than a tangled pipeline.

book_flight(flight_id=”AC-PHANTOM”, passenger_name=”J. Moffatt”)
# -> “Booked.” The action fired. Nothing in the loop asked whether it should.

Failure 2: context blows up and quality degrades silently

When working through the Failure 2 context blows stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Treat this stage as a contract between inputs and validated outputs. Name the artifacts, define success checks, and refuse silent partial completion. Log tool name, args hash, latency, and outcome for every call. Debugging agent loops without that trail wastes hours.

Failure 3: a tool errors, the agent reports success

When working through the Failure 3 a tool stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Record timings and token or query cost next to functional results. Cost visibility early prevents surprise bills when the path moves from demo to shared environments. Log tool name, args hash, latency, and outcome for every call. Debugging agent loops without that trail wastes hours.

The shape every part will follow

When working through the The shape every part stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Keep configuration outside application code. Environment files, secret stores, and feature flags belong in one place operators can audit without reading the whole graph. Log tool name, args hash, latency, and outcome for every call. Debugging agent loops without that trail wastes hours. When working through the The shape every part stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Prefer small, testable units over sprawling scripts. When a step fails, the failure should point at a single responsibility rather than a tangled pipeline.

Do this today

The Do this today stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Treat this stage as a contract between inputs and validated outputs. Name the artifacts, define success checks, and refuse silent partial completion. Expose tools with narrow schemas and explicit side-effect labels. Hosts need to know which calls mutate state before they auto-approve.

The model is the easy part

The The model is the stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Record timings and token or query cost next to functional results. Cost visibility early prevents surprise bills when the path moves from demo to shared environments. Budget tokens per turn and per session. Agentic tools expand context aggressively; hard caps keep demos from becoming surprise invoices.

Operational checklist

When working through the Operational checklist stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest.

Document the happy path and the recovery path together. Retries, human gates, and dead-letter handling are part of the product, not later polish.

Log tool name, args hash, latency, and outcome for every call. Debugging agent loops without that trail wastes hours.

Keep graph state flat and typed. Nested blobs hide which node wrote which field and break resume after interrupts.

Add a smoke test that exercises the critical path in CI with fixtures, not live paid APIs, whenever budgets allow.

Record timings and token or query cost next to functional results. Cost visibility early prevents surprise bills when the path moves from demo to shared environments.

Before promoting the stack, freeze versions, capture a golden transcript for the critical path, and confirm rollback steps. Shared environments need rate limits, tenancy checks, and a clear owner for secret rotation. Prefer boring reliability over clever one-off demos.

Batch note for 765280e2df21: keep provider keys out of the repo, set a per-session token ceiling, and store transcripts next to the eval fixtures so later model swaps stay comparable.