project-pitch

Files (62)

• LICENSE.txt
• SKILL.md
• frameworks.md · references/
• pptx-editing.md · references/
• pptx-skill.md · references/
• pptxgenjs.md · references/
• slide-guidelines.md · references/
• __init__.py · scripts/
• add_slide.py · scripts/
• clean.py · scripts/
• __init__.py · scripts/office/helpers/
• merge_runs.py · scripts/office/helpers/
• simplify_redlines.py · scripts/office/helpers/
• pack.py · scripts/office/
• dml-chart.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• dml-chartDrawing.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• dml-diagram.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• dml-lockedCanvas.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• dml-main.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• dml-picture.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• dml-spreadsheetDrawing.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• dml-wordprocessingDrawing.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• pml.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• shared-additionalCharacteristics.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• shared-bibliography.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• shared-commonSimpleTypes.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• shared-customXmlDataProperties.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• shared-customXmlSchemaProperties.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• shared-documentPropertiesCustom.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• shared-documentPropertiesExtended.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• shared-documentPropertiesVariantTypes.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• shared-math.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• shared-relationshipReference.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• sml.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• vml-main.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• vml-officeDrawing.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• vml-presentationDrawing.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• vml-spreadsheetDrawing.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• vml-wordprocessingDrawing.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• wml.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• xml.xsd · scripts/office/schemas/ISO-IEC29500-4_2016/
• opc-contentTypes.xsd · scripts/office/schemas/ecma/fouth-edition/
• opc-coreProperties.xsd · scripts/office/schemas/ecma/fouth-edition/
• opc-digSig.xsd · scripts/office/schemas/ecma/fouth-edition/
• opc-relationships.xsd · scripts/office/schemas/ecma/fouth-edition/
• mc.xsd · scripts/office/schemas/mce/
• wml-2010.xsd · scripts/office/schemas/microsoft/
• wml-2012.xsd · scripts/office/schemas/microsoft/
• wml-2018.xsd · scripts/office/schemas/microsoft/
• wml-cex-2018.xsd · scripts/office/schemas/microsoft/
• wml-cid-2016.xsd · scripts/office/schemas/microsoft/
• wml-sdtdatahash-2020.xsd · scripts/office/schemas/microsoft/
• wml-symex-2015.xsd · scripts/office/schemas/microsoft/
• soffice.py · scripts/office/
• unpack.py · scripts/office/
• validate.py · scripts/office/
• __init__.py · scripts/office/validators/
• base.py · scripts/office/validators/
• docx.py · scripts/office/validators/
• pptx.py · scripts/office/validators/
• …and 2 more

SKILL.md (first lines)

# Project Pitch Generator

You are a specialist in project pitch storytelling and presentation design. Your job is to produce:
1. A **fully timed narration script** for a pitch video
2. A **slide deck** — first as a reviewed markdown draft, then as an actual `.pptx` file

Work through the phases below in order. Do not skip or combine phases.

---

## PHASE 0 — CONTEXT DISCOVERY & INTAKE

### Step 0A — Scan the workspace

Before asking the user anything, scan the current workspace to auto-extract as much context as possible. Look for:
- `README.md`, `README.rst`, or any top-level markdown files
- Other `.md` files in the root or `/docs` folder
- `package.json`, `Cargo.toml`, `pyproject.toml`, `go.mod` — for tech stack
- Any `ARCHITECTURE.md`, `FEATURES.md`, or similar documentation
- Source code directories (for inferring tech stack if no manifest found)

From these, attempt to extract:
- **Project name & tagline**
- **What it does** (problem it solves, who uses it, what it produces)
- **Key features** (including any that are well-described in docs)
- **Tech stack** (languages, frameworks, key libraries)
- **Domain/industry** (e.g., "AI developer tooling", "healthcare")
- **Competitor/alternative tools** (if mentioned in docs)
- **Architectural "why" notes** (any design rationale in docs)
- **Brand phrases or terminology** (any recurring terms used as product-specific language)

### Step 0B — Present findings and ask for missing info

Present your findings to the user in this format:

---
**Here's what I found about your project. Please confirm, correct, or add to any of these:**

- **Project name & tagline:** [extracted or "Not found"]
- **What it does:** [extracted or "Not found"]
- **Key features:** [list or "Not found"]
- **Tech stack:** [extracted or "Not found"]
- **Domain:** [extracted or "Not found"]
- **Competitors/alternatives:** [extracted or "Not found"]
- **Brand phrases:** [extracted or "Not found"]
- **Architecture notes:** [extracted or "Not found"]

**I still need from you (required):**

1. **Demo flow** *(always required — cannot be inferred from code)*: Describe exactly what happens in your screen recording, step by step. Include what you do, what appears on screen, and the key "wow moments".

**Optional — leave blank to use defaults or skip:**

2. **Target audience**: Who is the audience? (e.g., technical evaluators, leadership, investors, mixed) — I'll assume a mixed technical/business audience if not specified.
3. **Time limit**: Total video length. Default: 3:00.
4. **Emotional tone**: e.g., urgent, cinematic, authoritative, energizing, conversational. Default: energizing + authoritative.
5. **Anything to add or correct** from the list above.

---

Wait for the user's response. If any of the required auto-extracted fields are still missing after the user's response, ask for them before proceeding.

---

## PHASE 1 — RESEARCH

Run web searches **concurrently** for:

1. **Industry stats** relevant to the domain:
   - Adoption numbers (how many people/companies use related tools)
   - Productivity data (time saved, time lost, costs)
   - Pain point statistics (developer surveys, analyst reports)
   - Market size / growth trajectory
   - Counterintuitive or surprising data that creates tension

2. **Competitor landscape** (supplement what user provided):
   - What existing alternatives do / don't do
   - Analyst framing of the gap (Gartner, Forrester, Stack Overflow Survey, etc.)

Collect the 5–8 most pitch-worthy stats. Prioritize:
- Recent (2024–2026)
- Citable sources (GitHub, Stack Overflow, Gartner, major analyst firms)
- Numbers with emotional weight (large %, specific time costs, surprising contrasts)

Do NOT show stats to the user yet — hold them for the framework outlines in Phase 2.

---

## PHASE 2 — FRAMEWORK SELECTION

Read [frameworks.md](references/frameworks.md) to understand the 5 storytelling frameworks.

For each framework, generate a **tailored rough outline** of 5–7 bullet points using the user's actual project, the stats from Phase 1, and the demo sequence. Make it specific enough that the user can picture the final script — not generic template language.

Present all 5 outlines:

```
## Framework 1: [Name]
**Best for:** [1-line assessment specific to this project]
**Rough outline:**
- [bullet — specific to this project, with a real stat woven in]
- ...

---
## Framework 2: [Name]
...
```

End with:
> **Recommendation:** [1–2 sentences on best fit and any hybrid suggestion]

Ask: *"Which framework would you like? You can also say 'hybrid' and specify what to combine."*

Wait for the user's response.

---

## PHASE 3 — NARRATION SCRIPT

From tezansahu/ai-garage, GPL-3.0.

More from tezansahu/ai-garage