board-speech-rules

Generates structured board-writing and speech-script execution tables for elementary math tutoring.

Updated Sep 6, 2026
One-click install
npx skills add https://github.com/4xiaxia/203 --skill board-speech-rules-4xiaxia
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: board-speech-rules
Source: https://github.com/4xiaxia/203/tree/main/skills/board-speech-rules
Command: npx skills add https://github.com/4xiaxia/203 --skill board-speech-rules-4xiaxia

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? It enforces consistent formatting rules when an AI agent generates five-field execution tables (stage, speech, board, actionSpec) for an elementary math tutoring system, preventing malformed coordinates, invalid LaTeX in speech text, and out-of-region drawing actions. ## Core Features & Use Cases - Board Writing Rules: Defines the v2.0 board object format with startCoord and content, fraction notation via \frac, and handwriting font symbol mappings. - Speech Script Rules: Provides pronunciation conversion tables for fractions, powers, roots, and compound expressions, plus banned TTS control tags and lecture-style phrases. - Action Tool Constraints: Restricts drawing to three registered tools (rough-notation, rough-line, rough-arrow) with region-bounded coordinates and triggerAt timing. - Use Case: When Agent B generates a JSON execution table for a fraction word problem, this Skill ensures the board shows \frac{7}{15} while the speech reads "15分之7", with coordinates matching the handoff boardPlan regions. ## Quick Start Use the board-speech-rules skill to generate a five-field execution table for this math problem using the provided handoff JSON.

Frequently Asked Questions about board-speech-rules

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate a board and speech execution table for math tutoring?

Provide the handoff JSON containing boardPlan, canvasParams, and zoneAnchors, then output rows with stage, speech, board, and actionSpec fields. Board uses the object format {startCoord, content}, and speech contains plain spoken text for TTS.

What coordinate format should board startCoord and actions use?

Coordinates follow the coordinateMode parameter: percentage mode uses 0-100 ranges with origin at top-left, while pixel mode uses canvasSize from handoff.canvasParams. All coordinates must stay within the corresponding region boundaries from boardPlan.

Which drawing tools are allowed in actionSpec?

Only three registered tools are permitted: rough-notation for underlines and highlights, rough-line for straight lines, and rough-arrow for arrows. Each action requires triggerAt timing, a unique order integer, and a region field.

Can speech output contain LaTeX or TTS control tags?

No. Speech must be plain spoken text without LaTeX commands, Markdown, or tags like [pause]. Fractions are spoken as denominator-first phrasing, and long pauses are achieved by splitting content into separate rows instead.

Why should the execution table not include duration or time fields?

Timing is computed dynamically by downstream audio and rendering systems based on actual parameters. The generator only outputs content fields; row gaps of about 1.5 seconds and writing speed jitter are handled by the renderer.