Practical notes: A Local Support Agent with SmolLM3: Think Mode, Tools, and

Use this as an operator-facing rebuild of the ideas in “A Local Support Agent with SmolLM3: Think Mode, Tools, and When Not to Reason”: clear stages, ordered code slots, and recovery notes that survive a handoff. The Overview 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.

The setup

For the The setup 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. Prefer structured outputs with schema validation over free-form prose when the next step is code or a tool call.

Think vs no_think is a product decision

For the Think vs nothink is 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. Prefer structured outputs with schema validation over free-form prose when the next step is code or a tool call.

messages = [
    {"role": "system", "content": "/no_think"},
    {"role": "user", "content": prompt},
]

tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True,
    enable_thinking=False,  # the /no_think flag in the system prompt wins if both are set
)

<|im_start|>assistant
<think>

</think>

final = re.sub(r"<think>.*?</think>", "", raw, flags=re.DOTALL).strip()

<tool_call>
{"name": <function-name>, "arguments": <args-json-object>}
</tool_call>

TOOLS = [
    {
        "name": "lookup_order_status",
        "description": (
            "Look up the current status, estimated delivery date, and carrier "
            "for a specific customer order. Call this when the customer mentions "
            "an order number or asks where their order is."
        ),
        "parameters": {
            "type": "object",
            "properties": {
                "order_id": {
                    "type": "string",
                    "description": "The order ID, usually in the format ORD-XXXXXX.",
                }
            },
            "required": ["order_id"],
        },
    }
]

ORDERS = {
    "ORD-4821": {"status": "shipped", "eta": "June 18, 2026", "carrier": "DHL"},
    "ORD-3307": {"status": "processing", "eta": "June 20, 2026", "carrier": None},
    "ORD-1190": {"status": "delivered", "eta": None, "carrier": "FedEx"},
}

def respond_with_tools(user_message: str) -> str:
    messages = [{"role": "user", "content": user_message}]
    turn1 = generate_reply(messages)
    name, args = parse_tool_call(turn1)
    if name == "lookup_order_status":
        result = lookup_order_status(**args)
        messages += [
            {"role": "assistant", "content": turn1},
            {"role": "tool", "content": json.dumps(result), "name": name},
        ]
        return generate_reply(messages)
    return turn1

<tool_call>
{"name": "lookup_order_status", "arguments": {"order_id": "ORD-4821"}}
</tool_call>

The email ticket is where the naive loop dies

For the The email ticket is 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. Prefer structured outputs with schema validation over free-form prose when the next step is code or a tool call. For the The email ticket is 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.

<tool_call>
{"name": "lookup_order_status", "arguments": {"order_id": "your_order_id"}}
</tool_call>

Guards that are not fine-tuning

When working through the Guards that are not 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. Cache stable system instructions and tool schemas. Re-sending identical preamble is a common source of burn.

ORDER_RE = re.compile(r"^ORD-\d{4,}quot;)
PURE_TOOL_RE = re.compile(r"^\s*<tool_call>(.*?)</tool_call>\s*quot;, re.DOTALL)

def parse_executable_tool_call(output: str):
    cleaned = re.sub(r"<think>.*?</think>", "", output, flags=re.DOTALL).strip()
    match = PURE_TOOL_RE.match(cleaned)
    if not match:
        return None
    payload = json.loads(match.group(1).strip())
    order_id = str(payload.get("arguments", {}).get("order_id", "")).strip()
    if not ORDER_RE.match(order_id):
        return None
    return payload

What this doesn’t prove

When working through the What this doesn t 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. Cache stable system instructions and tool schemas. Re-sending identical preamble is a common source of burn.

Run it

When working through the Run it 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. Cache stable system instructions and tool schemas. Re-sending identical preamble is a common source of burn. When working through the Run it 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.

python llm-learning/smollm3-local-support-agent/code/think_vs_nothink.py
python llm-learning/smollm3-local-support-agent/code/support_agent.py
python llm-learning/smollm3-local-support-agent/code/support_agent_guarded.py

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.

Prefer small, testable units over sprawling scripts. When a step fails, the failure should point at a single responsibility rather than a tangled pipeline.

Cache stable system instructions and tool schemas. Re-sending identical preamble is a common source of burn.

Expose tools with narrow schemas and explicit side-effect labels. Hosts need to know which calls mutate state before they auto-approve.

Put human approval on edges that spend money or change production data. Compile-time wiring does not equal business completeness.

Write a short runbook: how to rotate keys, how to drain the queue, how to roll back the last ingest.

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 0e44ad944ff5: 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. 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 0/868: 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. Treat this stage as a contract between inputs and validated outputs. Name the artifacts, define success checks, and refuse silent partial completion.

Hardening detail 1/868: 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. 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 2/868: 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. 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 3/868: 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.