read_patient

Reads a path of a patient's virtual file tree in a chosen format, optionally capped to a token budget.

Reads one path of a patient's virtual file tree, in the format and length you choose.

Input

Argument	Type	Required	Description
`patient_id`	string	Yes	The patient's `patientId`. See Shared arguments.
`path`	string	Yes	A path relative to the patient, such as `/labs/latest`. See Paths.
`format`	string	No	`narrative` (default), `structured` or `compact`. See Formats.
`token_budget`	number	No	Approximate cap on the output, in tokens of 4 characters. Omit for no cap.
`session_id`	string	No	A session to record the call on. See Record calls on a session.

Format	Output	Typical tokens
`narrative`	Prose Markdown with clinical context	300–800
`structured`	JSON with every field	200–500
`compact`	Key facts only	50–150

VFS Paths details what each format holds for each kind of path.

The content of the path, as text, in the requested format.

Input

{ "patient_id": "abc", "path": "/labs/latest", "format": "compact" }

Output

HbA1c 7.1% (2023-09-07) | eGFR 52 mL/min (2023-08-15) | ...

When the content is longer than token_budget × 4 characters, it is cut at that length and ends with …[truncated to token budget].

For a path that does not exist, Path not found: {path}. See Errors.