@bicharts/chart-mcp 0.6.101 → 0.6.103

package/README.md

@@ -63,6 +63,12 @@ The main tools:
   call signs you in) — no LLM is not the same as no cost, because eligibility *is* the
   engine's own answer. Free within a generous hourly allowance; beyond it a call may draw
   usage credits. A live license additionally unlocks *draft* types.
+- **`check_request`** — can this data honor the request as given? Takes `generate_chart`'s
+  arguments (the chart type and every setting) and answers, before anything is generated,
+  each ask this data can't honor and why, in the service's own words, with up to five types
+  that fit when the type itself is denied. **No LLM** and no generation credits; metered like
+  `list_eligible_charts`. On a service that doesn't offer the check, or has it switched off,
+  it says so and checks nothing.
 - **`suggest_charts`** — "which charts suit this model?" in one call, for a semantic model in a
   Microsoft Fabric App on Rayfin's data app template: reads the model's tables, measures and
   relationships through the project's own `fabric-app-data` CLI, finds its natural views (by place,
@@ -311,7 +317,7 @@ and `GET /healthz`. What's different from the local server:
   client signs in again.
 - **It's stateless:** each request gets a fresh server and no session, so `GET` and `DELETE` on `/mcp` answer 405,
   and a question a tool would ask the client (hover drop lines on a 3D scatter) goes to the agent instead.
-- **Five tools:** `assess_data_shape`, `list_eligible_charts`, `generate_chart`, `preview_request` and `get_guide`. The
+- **Six tools:** `assess_data_shape`, `list_eligible_charts`, `check_request`, `generate_chart`, `preview_request` and `get_guide`. The
   tools that work in your project folder (`check_chart`, `check_app`, `setup_fabric_app`, `query_semantic_model`,
   `suggest_charts`) need the local server, since a remote server can't see your machine.
 - **Tables travel in the request,** as `data` or `csv_text`; `csv_path`, `out_dir`, `preview_html` and `project_dir`