By name: weread-cognitive-shift description: "Generate a fixed-layout HTML 'cognitive shift' page from WeChat Reading data after the official WeChat Reading skill is already configured in the same environment. Use when the user wants to trace reading-driven changes in interests, thinking direction, and long-term cognitive threads: prompt for the user's WeChat Reading API key, collect reading data through the official skill, summarize it into the schema in references/cognitive_shift_schema.md, then render assets/cognitive-shift-template.html with scripts/render_cognitive_shift.py."
WeRead Cognitive Shift
Overview
Create a single-file HTML cognitive-shift page with a locked visual structure.
Keep the page layout, styles, and interactions in assets/cognitive-shift-template.html fixed by default. Only change the structured data payload unless the user explicitly asks to redesign the page.
Workflow
1. Confirm prerequisites
Confirm that the official WeChat Reading skill is already available in the same environment.
If it is missing, stop and tell the user to configure that skill first.
2. Collect only the required input
Ask for:
- the user's WeChat Reading
API Key - an optional page title only if the user wants to override the default
Do not ask the user to tune prompts or page structure by default.
3. Gather source data through the official skill
Collect the richest available evidence for:
- bookshelf / read books
- reading statistics
- highlights and notes
- representative books, authors, categories, keywords, and changes over time
Prefer higher-level official skill capabilities when they exist. If the official skill exposes lower-level endpoints instead, gather equivalent data and continue.
4. Build the structured payload
Read references/cognitive_shift_schema.md.
Use references/analysis_prompt.md as the default analysis prompt.
Produce JSON only. Do not generate HTML in this step.
5. Render the final page
Run:
python scripts/render_cognitive_shift.py --data <json-path> --output <html-path>
The renderer injects the JSON payload into assets/cognitive-shift-template.html and writes a standalone HTML file.
6. Return the deliverable
Return:
- the final HTML file path
- the JSON payload path if it is useful for debugging or later reuse
If the environment supports previewing local HTML, preview the result after rendering.
Rules
- Keep the HTML structure deterministic. Do not ask the model to freely rewrite the page.
- Treat the template as the source of truth for layout, CSS, and front-end behavior.
- Preserve all required schema keys even when some data must be inferred conservatively.
- Keep Chinese copy concise and evidence-backed. Avoid inflated prose.
- Prefer trend language that can be justified by reading stages, note density, or topic timing.
- Reduce noise. This page should focus on the clearest cognitive changes, not exhaustive book lists.
Resources
references/cognitive_shift_schema.md: output contract for the rendering payloadreferences/analysis_prompt.md: default constrained prompt for generating the payloadreferences/workbuddy_execution.md: runtime sequence for collecting data and rendering the pagescripts/render_cognitive_shift.py: deterministic rendererassets/cognitive-shift-template.html: final locked page template