This article is published in English.
Practical notes: Architecting the Agent Improvement Loop: Production-Grade Eval
Operable walkthrough of Practical notes: Architecting the Agent Improvement Loop: Production-Grade Eval: contracts, checks, and drop-in code slots for teams shipping this pattern.
The following notes reconstruct a practical path around “Architecting the Agent Improvement Loop: Production-Grade Eval Pipelines”. Emphasis stays on contracts, checks, and drop-in code placeholders rather than motivational framing.
Introduction
When working through the Introduction 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. Checkpoint after expensive steps. Resume should not re-bill the same LLM call when an operator retries a later node.
Table of Contents
When working through the Table of Contents 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. Checkpoint after expensive steps. Resume should not re-bill the same LLM call when an operator retries a later node.
Phase 1: Foundation
When working through the Phase 1 Foundation 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. Checkpoint after expensive steps. Resume should not re-bill the same LLM call when an operator retries a later node.
Setting up the environment
When working through the Setting up the environment 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. Checkpoint after expensive steps. Resume should not re-bill the same LLM call when an operator retries a later node.
# Install the libraries we need
pip install langgraph langchain langchain-openai langsmith
export LANGSMITH_TRACING=true
export LANGSMITH_API_KEY=your_langsmith_key
export LANGSMITH_PROJECT=agent-improvement-loop
export OPENAI_API_KEY=your_openai_key
# Step 1: Import the pieces we need
from langchain_openai import ChatOpenAI
from langchain_core.tools import tool
from langgraph.prebuilt import create_react_agent
# Step 2: Define two tools the agent can call
@tool
def get_account_info(account_id: str) -> str:
"""Look up basic information for an account by account_id."""
fake_accounts = {
"A100": "Account A100: plan=pro, status=active, email=ada@example.com",
"A200": "Account A200: plan=free, status=suspended, email=turing@example.com",
}
return fake_accounts.get(account_id, f"No account found with id {account_id}")
@tool
def get_recent_orders(account_id: str) -> str:
"""Return the three most recent orders for a given account_id."""
fake_orders = {
"A100": "Orders for A100: #9001 shipped, #9002 processing, #9003 returned",
"A200": "Orders for A200: no orders in the last 90 days",
}
return fake_orders.get(account_id, f"No orders for {account_id}")
# Step 3: Wire the tools into a ReAct agent
model = ChatOpenAI(model="gpt-4o-mini", temperature=0)
tools = [get_account_info, get_recent_orders]
agent = create_react_agent(model, tools)
Let’s test the agent
When working through the Let s test the 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. Checkpoint after expensive steps. Resume should not re-bill the same LLM call when an operator retries a later node.
# Step 4: Invoke the agent with a simple query
result = agent.invoke({
"messages": [{"role": "user", "content": "What plan is account A100 on?"}]
})
for message in result["messages"]:
print(message.type, ":", message.content)
When working through the Let s test the 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.
human : What plan is account A100 on?
ai :
tool : Account A100: plan=pro, status=active, email=ada@example.com
ai : Account A100 is on the pro plan.
Phase 2: Trace Collection
The Phase 2 Trace Collection 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. Keep graph state flat and typed. Nested blobs hide which node wrote which field and break resume after interrupts.
Pulling recent traces from production
The Pulling recent traces from 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. Keep graph state flat and typed. Nested blobs hide which node wrote which field and break resume after interrupts.
# Step 1: Create a LangSmith client
from langsmith import Client
from datetime import datetime, timedelta
client = Client()
# Step 2: List recent root runs from our project
recent_runs = list(client.list_runs(
project_name="agent-improvement-loop",
is_root=True,
start_time=datetime.utcnow() - timedelta(days=1),
))
print(f"Found {len(recent_runs)} root runs in the last 24 hours")
Found 42 root runs in the last 24 hours
Collecting the fields we need
The Collecting the fields we 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. Keep graph state flat and typed. Nested blobs hide which node wrote which field and break resume after interrupts. The Collecting the fields we 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.
# Step 3: Build a compact trace record from each run
def compact_trace(run):
return {
"run_id": str(run.id),
"inputs": run.inputs,
"outputs": run.outputs,
"error": run.error,
"latency_ms": (run.end_time - run.start_time).total_seconds() * 1000
if run.end_time else None,
"tool_calls": [
child.name for child in client.list_runs(
trace_id=run.trace_id, run_type="tool"
)
],
}
traces = [compact_trace(run) for run in recent_runs]
print(f"Collected {len(traces)} compact traces")
print("Example tool calls:", traces[0]["tool_calls"])
Collected 42 compact traces
Example tool calls: ['get_account_info']
# Step 4: Inspect the first trace end to end
import json
print(json.dumps(traces[0], indent=2, default=str))
{
"run_id": "b1d2e3f4-...",
"inputs": {"messages": [{"role": "user", "content": "What plan is account A100 on?"}]},
"outputs": {"messages": [{"role": "ai", "content": "Account A100 is on the pro plan."}]},
"error": null,
"latency_ms": 1423.51,
"tool_calls": ["get_account_info"]
}
Phase 3: Enriching Traces with Scores
For the Phase 3 Enriching Traces 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. Document the happy path and the recovery path together. Retries, human gates, and dead-letter handling are part of the product, not later polish. Put human approval on edges that spend money or change production data. Compile-time wiring does not equal business completeness.
Layer 1: Code based tool correctness check
For the Layer 1 Code based 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. Authenticate at the gateway and re-authorize at the data plane. A bearer token alone is not a tenancy boundary.
# Step 1: Define a rule that says which tool SHOULD be called for a given query
import re
def expected_tool_for_query(query: str) -> str:
q = query.lower()
if re.search(r"\b(order|shipment|shipped|return)\b", q):
return "get_recent_orders"
if re.search(r"\b(plan|account|status|email)\b", q):
return "get_account_info"
return "any"
def score_tool_correctness(trace):
query = trace["inputs"]["messages"][0]["content"]
expected = expected_tool_for_query(query)
actual = trace["tool_calls"][0] if trace["tool_calls"] else None
if expected == "any":
return 1.0
return 1.0 if actual == expected else 0.0
# Step 2: Score qualitative properties with an LLM judge
from langchain_openai import ChatOpenAI
from langchain_core.prompts import ChatPromptTemplate
judge_model = ChatOpenAI(model="gpt-4o-mini", temperature=0)
judge_prompt = ChatPromptTemplate.from_messages([
("system", "You are an expert reviewer of AI agent responses. "
"Rate the response on a scale of 1 to 5 for helpfulness. "
"Respond with only a single integer, nothing else."),
("user", "User question: {question}\n\nAgent response: {response}\n\nScore:"),
])
def score_helpfulness(trace):
question = trace["inputs"]["messages"][0]["content"]
response = trace["outputs"]["messages"][-1]["content"]
chain = judge_prompt | judge_model
result = chain.invoke({"question": question, "response": response})
try:
return int(result.content.strip()) / 5.0
except ValueError:
return 0.0
Layer 3: the author review as an annotation queue
For the Layer 3 the author review 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. Put human approval on edges that spend money or change production data. Compile-time wiring does not equal business completeness. For the Layer 3 the author review 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.
# Step 3: Model a human review signal
def score_human(trace, human_labels: dict) -> float | None:
run_id = trace["run_id"]
return human_labels.get(run_id)
human_labels = {
"b1d2e3f4-...": 1.0,
"c2e3f4g5-...": 0.0,
}
Combining the three layers
When working through the Combining the three layers 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. Checkpoint after expensive steps. Resume should not re-bill the same LLM call when an operator retries a later node.
# Step 4: Run all three scorers and attach scores to each trace
def enrich(trace):
trace["scores"] = {
"tool_correctness": score_tool_correctness(trace),
"helpfulness": score_helpfulness(trace),
"human": score_human(trace, human_labels),
}
return trace
enriched = [enrich(t) for t in traces]
print(json.dumps(enriched[0]["scores"], indent=2))
{
"tool_correctness": 1.0,
"helpfulness": 0.8,
"human": null
}
Phase 4: Pattern Discovery
When working through the Phase 4 Pattern Discovery 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. Checkpoint after expensive steps. Resume should not re-bill the same LLM call when an operator retries a later node.
Filtering for failures
When working through the Filtering for failures 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. Checkpoint after expensive steps. Resume should not re-bill the same LLM call when an operator retries a later node.
# Step 1: Keep only traces where at least one score is low
def is_low_scoring(trace):
scores = trace["scores"]
if scores["human"] is not None and scores["human"] < 0.5:
return True
if scores["tool_correctness"] < 0.5:
return True
if scores["helpfulness"] < 0.6:
return True
return False
failures = [t for t in enriched if is_low_scoring(t)]
print(f"{len(failures)} of {len(enriched)} traces flagged as low scoring")
When working through the Filtering for failures 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.
9 of 42 traces flagged as low scoring
Clustering failures into categories
The Clustering failures into categories 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. Keep graph state flat and typed. Nested blobs hide which node wrote which field and break resume after interrupts.
# Step 2: Tag each failure with a category based on its scores
def categorize(trace):
scores = trace["scores"]
if scores["tool_correctness"] < 0.5:
return "wrong_tool"
if scores["helpfulness"] < 0.6 and scores["tool_correctness"] >= 0.5:
return "unhelpful_answer"
if scores["human"] is not None and scores["human"] < 0.5:
return "human_flagged"
return "other"
from collections import Counter
categories = Counter(categorize(t) for t in failures)
print(categories.most_common())
[('wrong_tool', 5), ('unhelpful_answer', 3), ('human_flagged', 1)]
# Step 3: Print the inputs and outputs for every wrong_tool failure
wrong_tool = [t for t in failures if categorize(t) == "wrong_tool"]
for trace in wrong_tool[:3]:
print("QUERY:", trace["inputs"]["messages"][0]["content"])
print("TOOLS CALLED:", trace["tool_calls"])
print("RESPONSE:", trace["outputs"]["messages"][-1]["content"])
print("---")
QUERY: Is my subscription active?
TOOLS CALLED: ['get_recent_orders']
RESPONSE: I could not find any recent orders to confirm your subscription status.
---
QUERY: Tell me about my subscription to your product
TOOLS CALLED: ['get_recent_orders']
RESPONSE: There are no recent orders for this account.
---
Phase 5: Offline Test Suites
The Phase 5 Offline Test 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. Keep graph state flat and typed. Nested blobs hide which node wrote which field and break resume after interrupts.
Creating the dataset
The Creating the dataset 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. Keep graph state flat and typed. Nested blobs hide which node wrote which field and break resume after interrupts.
# Step 1: Create a LangSmith dataset for our agent
dataset_name = "agent-production-failures"
dataset = client.create_dataset(
dataset_name=dataset_name,
description="Production traces where the agent failed. Each example captures the user query and the expected tool call.",
)
print(f"Created dataset: {dataset.id}")
The Creating the dataset 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.
Created dataset: 3f2a1b0c-...
# Step 2: Convert each failure trace into a dataset example
examples = []
for trace in failures:
query = trace["inputs"]["messages"][0]["content"]
expected_tool = expected_tool_for_query(query)
examples.append({
"inputs": {"question": query},
"outputs": {"expected_tool": expected_tool},
"metadata": {"category": categorize(trace), "source_run_id": trace["run_id"]},
})
client.create_examples(dataset_id=dataset.id, examples=examples)
print(f"Added {len(examples)} examples to {dataset_name}")
Added 9 examples to agent-production-failures
# Step 3: Write an evaluator that checks the agent's tool call against the expected tool
def tool_match_evaluator(inputs: dict, outputs: dict, reference_outputs: dict):
actual_messages = outputs.get("messages", [])
tool_called = None
for m in actual_messages:
if getattr(m, "tool_calls", None):
tool_called = m.tool_calls[0]["name"]
break
expected = reference_outputs["expected_tool"]
score = 1.0 if (expected == "any" or tool_called == expected) else 0.0
return {"key": "tool_match", "score": score}
The target function
For the The target function 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. Document the happy path and the recovery path together. Retries, human gates, and dead-letter handling are part of the product, not later polish. Put human approval on edges that spend money or change production data. Compile-time wiring does not equal business completeness.
# Step 4: Wrap our agent in a function that takes a LangSmith example and returns its output
def run_agent(inputs: dict) -> dict:
result = agent.invoke({
"messages": [{"role": "user", "content": inputs["question"]}]
})
return result
Phase 6: Closing the Loop
For the Phase 6 Closing the 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. Put human approval on edges that spend money or change production data. Compile-time wiring does not equal business completeness.
Running the evaluation
For the Running the evaluation 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. Put human approval on edges that spend money or change production data. Compile-time wiring does not equal business completeness.
# Step 1: Run the evaluator across the dataset with the current agent
from langsmith.evaluation import evaluate
experiment_results = evaluate(
run_agent,
data=dataset_name,
evaluators=[tool_match_evaluator],
experiment_prefix="agent-v1-baseline",
max_concurrency=4,
)
For the Running the evaluation 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.
View the evaluation results for experiment: 'agent-v1-baseline-...' at:
https://smith.langchain.com/o/.../datasets/.../compare?selectedSessions=...
9/9 runs | avg tool_match: 0.33
Shipping a fix and re-evaluating
When working through the Shipping a fix and 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. Checkpoint after expensive steps. Resume should not re-bill the same LLM call when an operator retries a later node.
# Step 2: Update the tool docstring to teach the agent about subscriptions
@tool
def get_account_info_v2(account_id: str) -> str:
"""Look up basic information for an account, including plan, status,
subscription tier, and contact email. Use this tool for any question
about the account itself, including subscription status."""
fake_accounts = {
"A100": "Account A100: plan=pro, status=active, email=ada@example.com",
"A200": "Account A200: plan=free, status=suspended, email=turing@example.com",
}
return fake_accounts.get(account_id, f"No account found with id {account_id}")
tools_v2 = [get_account_info_v2, get_recent_orders]
agent_v2 = create_react_agent(model, tools_v2)
def run_agent_v2(inputs: dict) -> dict:
result = agent_v2.invoke({
"messages": [{"role": "user", "content": inputs["question"]}]
})
return result
# Step 3: Run the evaluator on the v2 agent
experiment_results_v2 = evaluate(
run_agent_v2,
data=dataset_name,
evaluators=[tool_match_evaluator],
experiment_prefix="agent-v2-docstring-fix",
max_concurrency=4,
)
View the evaluation results for experiment: 'agent-v2-docstring-fix-...' at:
https://smith.langchain.com/o/.../datasets/.../compare?selectedSessions=...
9/9 runs | avg tool_match: 0.89
Before and after
When working through the Before and after 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. Checkpoint after expensive steps. Resume should not re-bill the same LLM call when an operator retries a later node.
Before (v1): After (v2):
wrong_tool: 5 wrong_tool: 1
unhelpful: 3 unhelpful: 0
human: 1 human: 0
avg score: 0.33 avg score: 0.89
Evaluation and Testing
When working through the Evaluation and Testing 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. Checkpoint after expensive steps. Resume should not re-bill the same LLM call when an operator retries a later node. When working through the Evaluation and Testing 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.
# Full loop driver
def run_improvement_loop():
# 1. Pull traces
runs = list(client.list_runs(
project_name="agent-improvement-loop",
is_root=True,
start_time=datetime.utcnow() - timedelta(days=1),
))
traces = [compact_trace(r) for r in runs]
# 2. Enrich with scores
enriched = [enrich(t) for t in traces]
# 3. Find failures and categorize them
failures = [t for t in enriched if is_low_scoring(t)]
categories = Counter(categorize(t) for t in failures)
# 4. Add failures to the dataset
new_examples = [{
"inputs": {"question": t["inputs"]["messages"][0]["content"]},
"outputs": {"expected_tool": expected_tool_for_query(
t["inputs"]["messages"][0]["content"])},
"metadata": {"category": categorize(t), "source_run_id": t["run_id"]},
} for t in failures]
if new_examples:
client.create_examples(dataset_id=dataset.id, examples=new_examples)
# 5. Re-evaluate the current agent against the full dataset
result = evaluate(
run_agent_v2,
data=dataset_name,
evaluators=[tool_match_evaluator],
experiment_prefix="daily-regression",
)
return {
"traces_collected": len(traces),
"failures_found": len(failures),
"by_category": dict(categories),
"examples_added": len(new_examples),
}
print(run_improvement_loop())
{
"traces_collected": 186,
"failures_found": 12,
"by_category": {"wrong_tool": 4, "unhelpful_answer": 6, "human_flagged": 2},
"examples_added": 12,
"regression_score": 0.87
}
How to Improve It Further
The How to Improve It 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. Keep graph state flat and typed. Nested blobs hide which node wrote which field and break resume after interrupts.
Operational checklist
For the Operational checklist 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.
Put human approval on edges that spend money or change production data. Compile-time wiring does not equal business completeness.
Track cost and latency beside quality. A slightly worse answer that costs 10x less may be the right production trade.
Pin dependency versions and record the image digest that ran the demo. Reproducibility beats tribal knowledge.
Prefer small, testable units over sprawling scripts. When a step fails, the failure should point at a single responsibility rather than a tangled pipeline.
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 febbeae44915: 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.
When working through the hardening note 0 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.
Hardening detail 0/961: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.
The hardening note 1 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.
Hardening detail 1/961: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.
For the hardening note 2 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. Document the happy path and the recovery path together. Retries, human gates, and dead-letter handling are part of the product, not later polish.
Hardening detail 2/961: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.
When working through the hardening note 3 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.
Hardening detail 3/961: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.
The hardening note 4 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.
Hardening detail 4/961: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.
For the hardening note 5 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.
Hardening detail 5/961: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.
When working through the hardening note 6 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.
Hardening detail 6/961: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.
The hardening note 7 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.
Hardening detail 7/961: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.
For the hardening note 8 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.
Hardening detail 8/961: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.
When working through the hardening note 9 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.
Hardening detail 9/961: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.
The hardening note 10 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.
Hardening detail 10/961: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.