PPT Master Skill

AI-driven multi-format SVG content generation system. Converts source documents into high-quality SVG pages through multi-role collaboration and exports to PPTX.

Core Pipeline: Source Document → Create Project → Template Option → Strategist → [Image_Generator] → Executor → Post-processing → Export

[!CAUTION]

🚨 Global Execution Discipline (MANDATORY)

This workflow is a strict serial pipeline. The following rules have the highest priority — violating any one of them constitutes execution failure:

1. SERIAL EXECUTION — Steps MUST be executed in order; the output of each step is the input for the next. Non-BLOCKING adjacent steps may proceed continuously once prerequisites are met, without waiting for the user to say "continue"
2. BLOCKING = HARD STOP — Steps marked ⛔ BLOCKING require a full stop; the AI MUST wait for an explicit user response before proceeding and MUST NOT make any decisions on behalf of the user
3. NO CROSS-PHASE BUNDLING — Cross-phase bundling is FORBIDDEN. (Note: the Eight Confirmations in Step 4 are ⛔ BLOCKING — the AI MUST present recommendations and wait for explicit user confirmation before proceeding. Once the user confirms, all subsequent non-BLOCKING steps — design spec output, SVG generation, speaker notes, and post-processing — may proceed automatically without further user confirmation)
4. GATE BEFORE ENTRY — Each Step has prerequisites (🚧 GATE) listed at the top; these MUST be verified before starting that Step
5. NO SPECULATIVE EXECUTION — "Pre-preparing" content for subsequent Steps is FORBIDDEN (e.g., writing SVG code during the Strategist phase)
6. NO SUB-AGENT SVG GENERATION — Executor Step 6 SVG generation is context-dependent and MUST be completed by the current main agent end-to-end. Delegating page SVG generation to sub-agents is FORBIDDEN
7. SEQUENTIAL PAGE GENERATION ONLY — In Executor Step 6, after the global design context is confirmed, SVG pages MUST be generated sequentially page by page in one continuous pass. Grouped page batches (for example, 5 pages at a time) are FORBIDDEN
8. SPEC_LOCK RE-READ PER PAGE — Before generating each SVG page, Executor MUST read_file <project_path>/spec_lock.md. All colors / fonts / icons / images MUST come from this file — no values from memory or invented on the fly. Executor MUST also look up the current page's page_rhythm tag and apply the matching layout discipline (anchor / dense / breathing — see executor-base.md §2.1). This rule exists to resist context-compression drift on long decks and to break the uniform "every page is a card grid" default

[!IMPORTANT]

🌐 Language & Communication Rule

• Response language: Always match the language of the user's input and provided source materials. For example, if the user asks in Chinese, respond in Chinese; if the source material is in English, respond in English.
• Explicit override: If the user explicitly requests a specific language (e.g., "请用英文回答" or "Reply in Chinese"), use that language instead.
• Template format: The design_spec.md file MUST always follow its original English template structure (section headings, field names), regardless of the conversation language. Content values within the template may be in the user's language.

[!IMPORTANT]

🔌 Compatibility With Generic Coding Skills

• ppt-master is a repository-specific workflow skill, not a general application scaffold
• Do NOT create or require .worktrees/, tests/, branch workflows, or other generic engineering structure by default
• If another generic coding skill suggests repository conventions that conflict with this workflow, follow this skill first unless the user explicitly asks otherwise

Main Pipeline Scripts

For complete tool documentation, see ${SKILL_DIR}/scripts/README.md.

Template Index

Standalone Workflows

Workflow

Step 1: Source Content Processing

🚧 GATE: User has provided source material (PDF / DOCX / EPUB / URL / Markdown file / text description / conversation content — any form is acceptable).

When the user provides non-Markdown content, convert immediately:

✅ Checkpoint — Confirm source content is ready, proceed to Step 2.

Step 2: Project Initialization

🚧 GATE: Step 1 complete; source content is ready (Markdown file, user-provided text, or requirements described in conversation are all valid).

python3 ${SKILL_DIR}/scripts/project_manager.py init <project_name> --format <format>

Format options: ppt169 (default), ppt43, xhs, story, etc. For the full format list, see references/canvas-formats.md.

Import source content (choose based on the situation):

⚠️ MUST use --move: All source files (original PDF / MD / images) MUST be moved (not copied) into sources/ for archiving.

• Markdown files generated in Step 1, original PDFs, original MDs — all must be moved into the project via import-sources --move
• Intermediate artifacts (e.g., _files/ directories) are handled automatically by import-sources
• After execution, source files no longer exist at their original location

✅ Checkpoint — Confirm project structure created successfully, sources/ contains all source files, converted materials are ready. Proceed to Step 3.

Step 3: Template Option

🚧 GATE: Step 2 complete; project directory structure is ready.

Default path — free design, no question asked. Proceed directly to Step 4. Do NOT query layouts_index.json and do NOT ask the user an A/B template-vs-free-design question. Free design is the standard mode: the AI tailors structure and style to the specific content.

Template flow is opt-in. Enter it only when one of these explicit triggers appears in the user's prior messages:

1. User names a specific template (e.g., "用 mckinsey 模板" / "use the academic_defense template")
2. User names a style / brand reference that maps to a template (e.g., "McKinsey 那种" / "Google style" / "学术答辩样式")
3. User explicitly asks what templates exist (e.g., "有哪些模板可以用")

Only when a trigger fires: read ${SKILL_DIR}/templates/layouts/layouts_index.json, resolve the match (or list available options for trigger 3), and copy template files to the project directory:

cp ${SKILL_DIR}/templates/layouts/<template_name>/*.svg <project_path>/templates/
cp ${SKILL_DIR}/templates/layouts/<template_name>/design_spec.md <project_path>/templates/
cp ${SKILL_DIR}/templates/layouts/<template_name>/*.png <project_path>/images/ 2>/dev/null || true
cp ${SKILL_DIR}/templates/layouts/<template_name>/*.jpg <project_path>/images/ 2>/dev/null || true

Soft hint (non-blocking, optional). Before Step 4, if the user's content is an obvious strong match for an existing template (e.g., clearly an academic defense, a government report, a McKinsey-style consulting deck) AND the user has given no template signal, the AI MAY emit a single-sentence notice and continue without waiting:

Note: the library has a template <name> that matches this scenario closely. Say the word if you want to use it; otherwise I'll continue with free design.

This is a hint, not a question — do NOT block, do NOT require an answer. Skip the hint entirely when the match is weak or ambiguous.

To create a new global template, read workflows/create-template.md

✅ Checkpoint — Default path proceeds to Step 4 without user interaction. If a template trigger fired, template files are copied before advancing.

Step 4: Strategist Phase (MANDATORY — cannot be skipped)

🚧 GATE: Step 3 complete; default free-design path taken, or (if triggered) template files copied into the project.

First, read the role definition:

Read references/strategist.md

⚠️ Mandatory gate in strategist.md: Before writing design_spec.md, Strategist MUST read_file templates/design_spec_reference.md and produce the spec following its full I–XI section structure. See strategist.md Section 1 for the explicit gate rule.

Must complete the Eight Confirmations (full template structure in templates/design_spec_reference.md):

⛔ BLOCKING: The Eight Confirmations MUST be presented to the user as a bundled set of recommendations, and you MUST wait for the user to confirm or modify before outputting the Design Specification & Content Outline. This is the single core confirmation point in the workflow. Once confirmed, all subsequent script execution and slide generation should proceed fully automatically.

1. Canvas format
2. Page count range
3. Target audience
4. Style objective
5. Color scheme
6. Icon usage approach
7. Typography plan
8. Image usage approach

If the user has provided images, run the analysis script before outputting the design spec (do NOT directly read/open image files — use the script output only):

python3 ${SKILL_DIR}/scripts/analyze_images.py <project_path>/images

⚠️ Image handling rule: The AI must NEVER directly read, open, or view image files (.jpg, .png, etc.). All image information must come from the analyze_images.py script output or the Design Specification's Image Resource List.

Output:

• <project_path>/design_spec.md — human-readable design narrative
• <project_path>/spec_lock.md — machine-readable execution contract (distilled from the decisions in design_spec.md; Executor re-reads this before every page). See templates/spec_lock_reference.md for the skeleton.

✅ Checkpoint — Phase deliverables complete, auto-proceed to next step:

## ✅ Strategist Phase Complete
- [x] Eight Confirmations completed (user confirmed)
- [x] Design Specification & Content Outline generated
- [x] Execution lock (spec_lock.md) generated
- [ ] **Next**: Auto-proceed to [Image_Generator / Executor] phase

Step 5: Image_Generator Phase (Conditional)

🚧 GATE: Step 4 complete; Design Specification & Content Outline generated and user confirmed.

Trigger condition: Image approach includes "AI generation". If not triggered, skip directly to Step 6 (Step 6 GATE must still be satisfied).

Read references/image-generator.md

1. Extract all images with status "pending generation" from the design spec
2. Generate prompt document → <project_path>/images/image_prompts.md
3. Generate images (CLI tool recommended):

   python3 ${SKILL_DIR}/scripts/image_gen.py "prompt" --aspect_ratio 16:9 --image_size 1K -o <project_path>/images
   

✅ Checkpoint — Confirm image generation attempted for every row, proceed to Step 6:

## ✅ Image_Generator Phase Complete
- [x] Prompt document created
- [x] Each image: status is either `Generated` (file present in images/) or `Needs-Manual` (reported to user with filename + reason)
- [x] No row remains `Pending`

On generation failure, do NOT halt — follow the Failure Handling rule in references/image-generator.md §4.3: retry once, then mark the row Needs-Manual, report to user, and continue to Step 6.

Step 6: Executor Phase

🚧 GATE: Step 4 (and Step 5 if triggered) complete; all prerequisite deliverables are ready.

Read the role definition based on the selected style:

Read references/executor-base.md          # REQUIRED: common guidelines
Read references/executor-general.md       # General flexible style
Read references/executor-consultant.md    # Consulting style
Read references/executor-consultant-top.md # Top consulting style (MBB level)

Only need to read executor-base + one style file.

Design Parameter Confirmation (Mandatory): Before generating the first SVG, the Executor MUST review and output key design parameters from the Design Specification (canvas dimensions, color scheme, font plan, body font size) to ensure spec adherence. See executor-base.md Section 2 for details.

Per-page spec_lock re-read (Mandatory): Before generating each SVG page, Executor MUST read_file <project_path>/spec_lock.md and use only the colors / fonts / icons / images listed there. This resists context-compression drift on long decks. See executor-base.md §2.1 for details.

⚠️ Main-agent only rule: SVG generation in Step 6 MUST remain with the current main agent because page design depends on full upstream context (source content, design spec, template mapping, image decisions, and cross-page consistency). Do NOT delegate any slide SVG generation to sub-agents. ⚠️ Generation rhythm rule: After confirming the global design parameters, the Executor MUST generate pages sequentially, one page at a time, while staying in the same continuous main-agent context. Do NOT split Step 6 into grouped page batches such as 5 pages per batch.

Visual Construction Phase:

• Generate SVG pages sequentially, one page at a time, in one continuous pass → <project_path>/svg_output/

Quality Check Gate (Mandatory) — after all SVGs are generated and BEFORE speaker notes:

python3 ${SKILL_DIR}/scripts/svg_quality_checker.py <project_path>

• Any error (banned SVG features, viewBox mismatch, spec_lock drift, etc.) MUST be fixed on the offending page before proceeding — go back to Visual Construction, re-generate that page, re-run the check.
• warning entries (e.g., low-resolution image, non-PPT-safe font tail) should be reviewed and fixed when straightforward; may be acknowledged and released otherwise.
• Running the checker against svg_output/ is required — running it only after finalize_svg.py is too late (finalize rewrites SVG and some violations get masked).

Logic Construction Phase:

• Generate speaker notes → <project_path>/notes/total.md

✅ Checkpoint — Confirm all SVGs and notes are fully generated and quality-checked. Proceed directly to Step 7 post-processing:

## ✅ Executor Phase Complete
- [x] All SVGs generated to svg_output/
- [x] svg_quality_checker.py passed (0 errors)
- [x] Speaker notes generated at notes/total.md

Step 7: Post-processing & Export

🚧 GATE: Step 6 complete; all SVGs generated to svg_output/; speaker notes notes/total.md generated.

⚠️ The following three sub-steps MUST be executed individually one at a time. Each command must complete and be confirmed successful before running the next. ❌ NEVER put all three commands in a single code block or single shell invocation.

Step 7.1 — Split speaker notes:

python3 ${SKILL_DIR}/scripts/total_md_split.py <project_path>

Step 7.2 — SVG post-processing (icon embedding / image crop & embed / text flattening / rounded rect to path):

python3 ${SKILL_DIR}/scripts/finalize_svg.py <project_path>

Step 7.3 — Export PPTX (embeds speaker notes by default):

python3 ${SKILL_DIR}/scripts/svg_to_pptx.py <project_path> -s final
# Output: exports/<project_name>_<timestamp>.pptx + exports/<project_name>_<timestamp>_svg.pptx
# Use --only native  to skip SVG reference version
# Use --only legacy  to only generate SVG image version

❌ NEVER use cp as a substitute for finalize_svg.py — it performs multiple critical processing steps ❌ NEVER export directly from svg_output/ — MUST use -s final to export from svg_final/ ❌ NEVER add extra flags like --only

Role Switching Protocol

Before switching roles, you MUST first read the corresponding reference file — skipping is FORBIDDEN. Output marker:

## [Role Switch: <Role Name>]
📖 Reading role definition: references/<filename>.md
📋 Current task: <brief description>

Reference Resources

Notes

• Do NOT add extra flags like --only to the post-processing commands — run them as-is
• Local preview: python3 -m http.server -d <project_path>/svg_final 8000
• Troubleshooting: If the user encounters issues during generation (layout overflow, export errors, blank images, etc.), recommend checking docs/faq.md — it contains known solutions sourced from real user reports and is continuously updated