Mineral Process Flowchart Generation

What It Addresses

Mineral processing test reports often contain flow descriptions such as crushing, grinding, flotation, and gravity separation, together with yield, grade, and recovery values. Drawing mass-flow diagrams by hand requires repeated material balance checks, and text, data, and final drawings can drift apart. This skill targets digitizing mineral processing reports and drawing process flowcharts, converting unstructured reports into reviewable process topology, corrected material-flow indicators, and a professionally labeled engineering flowchart with Simplified Chinese annotations.

How It Works

The skill follows an engineering workflow of parsing, balancing, rendering, and delivery:
- Compliance check: validates the report text and assay data before large-model processing to reduce unstable inputs.
- Topology extraction and balance calculation in parallel: identifies operating nodes such as roughing, cleaning, and scavenging, then uses methods such as least squares to correct and balance material flows using yield and grade data.
- Prompt fusion and image generation: combines structured topology and balanced data into detailed text-to-image prompts, then generates a 3:4 process flowchart.
- Async polling and delivery: real invocations submit a task, return a task_id, and poll until the image URL, node/edge data, corrected yield, grade, and recovery values are available.

Boundaries and Notes

report_text should include a reasonably complete mineral processing test report; adding assay_data as standard JSON grade data can improve balance accuracy. The skill is required to run asynchronously rather than synchronously. Without showapi_appkey, it triggers a simulated call for preview only, not real generation.