Drill · lap 2 · 3/10 · TOOL-102 · Catch a hallucinated argument before it crashes the tool

[akash-coded]

3 of 10 · TOOL-102 · Catch a hallucinated argument before it crashes the tool

tools · medium · implement · ~12 min · no AWS account

Models invent parameter names. booking_reference instead of booking_ref. flight instead of flight_no. Left alone, that is a TypeError from **kwargs — the single most common dispatch crash in production. Validate first, and answer in a way the model can act on.

def validate_args(schema: dict, args: dict | None) -> dict:
    """schema = {"required": ["booking_ref"], "properties": {"booking_ref": "string", "verbose": "boolean"}}

    Returns {"ok": True} when every supplied key is known and every required key is present.
    Otherwise {"ok": False,
               "unknown": [...],          # keys the schema does not declare
               "missing": [...],          # required keys that are absent
               "advice": "..."}           # names the VALID keys, so the model can retry correctly
    Never raises — args may be None.
    """

/drill TOOL-102

```python
def validate_args(schema, args):
    ...
```

What this proves

That you treat model-supplied arguments as untrusted input, and that your error message does the one thing that lets the model self-correct: it says what would have been valid.

---

Submit in this thread

Copy the block from the brief, edit it, and post it as a comment here. The bot replies within a couple of minutes with every check, the skill you demonstrated, and what to do next.

_{Brief in the repo: labs/drills/tools/TOOL-102 · next in the sequence: MEM-102 · how grading works · scoreboard}