pitch-deck-reviewer

📁 kaustav-payday/pitch-deck-reviewer 📅 1 day ago
2
总安装量
1
周安装量
#73897
全站排名
安装命令
npx skills add https://github.com/kaustav-payday/pitch-deck-reviewer --skill pitch-deck-reviewer

Agent 安装分布

amp 1
cline 1
opencode 1
cursor 1
kimi-cli 1
codex 1

Skill 文档

Pitch Deck Reviewer

Review pitch decks through the lens of two complementary frameworks. Read references/frameworks.md before starting any review.

Review Workflow

  1. Read the deck — Go through every slide, extracting the core message of each.
  2. Map slides to the 7 Blocks — Identify which Block(s) each slide addresses. Flag slides that don’t map to any Block.
  3. Evaluate each slide on the ABC dual test:
    • Does this slide reduce risk for the investor (making A→C more credible)?
    • Does this slide increase the value of C (making the outcome bigger/more defensible)?
    • Rate each: Strong / Weak / Missing
  4. Assess the 4 Conviction Layers — For each layer (Market, Insight, Founder, Execution), rate how well the deck builds conviction: Strong / Needs Work / Missing.
  5. Check narrative flow — Do the slides follow the Inevitable Story arc (Blocks 1→7)? Flag breaks in logic or missing narrative bridges.
  6. Audit deck economy — Identify bloat and consolidation opportunities:
    • Slides covering the same Block that should be merged into one punchy slide
    • Slides that dilute a point by spreading it too thin across multiple slides
    • Filler slides (appendix-style data, generic market size charts, logo walls) that add length without conviction
    • Overall slide count vs. the ideal 10–15 range for a VC deck
  7. Produce the review output.

Review Output Format

Slide-by-Slide Analysis

For each slide:

Slide N: [Slide Title/Topic]

  • Block: [Which of the 7 Blocks this maps to, or “None — consider removing/reworking”]
  • Conviction Layer: [MARKET / INSIGHT / FOUNDER / EXECUTION]
  • Reduces Risk? [What specific risk it reduces, or “No — here’s why”]
  • Increases Value of C? [How it expands the future outcome, or “No — here’s why”]
  • Verdict: [Strong / Needs Work / Weak]
  • Suggestion: [One concrete improvement if needed]

Narrative Arc Assessment

Evaluate how well the deck follows the 7-Block sequence. Identify:

  • Missing blocks (gaps in the story)
  • Out-of-order blocks (narrative confusion)
  • Blocks that are present but unconvincing

Conviction Layer Scorecard

Layer Rating Key Gap
MARKET Strong / Needs Work / Missing [Specific gap]
INSIGHT Strong / Needs Work / Missing [Specific gap]
FOUNDER Strong / Needs Work / Missing [Specific gap]
EXECUTION Strong / Needs Work / Missing [Specific gap]

ABC Framework Summary

  • Point A (Current State): [How well the deck establishes where they are today]
  • Point C (Future Vision): [How compelling and large C feels]
  • A→C Journey: [Does it feel clear, credible, and inevitable?]
  • Valuation Gap: [Does the current stage (A) suggest the price is meaningfully below C?]

Deck Economy

  • Slide count: [N slides] — [Over / Under / Within] the 10–15 ideal range
  • Merge candidates: List slides that cover the same Block and would land harder as a single slide. Explain what the merged slide should say.
  • Cut candidates: List slides that add length without adding conviction. Explain why.
  • Split candidates (rare): List slides trying to do too much that should be broken apart.

Top 3 Improvements

Prioritized, actionable changes that would most improve investor conviction.

Evaluation Principles

  • A slide that neither reduces risk nor increases C is wasted space — flag it.
  • The strongest slides do both simultaneously.
  • Block 4 (Earned Secret) is the hardest to nail and the most differentiated — pay extra attention.
  • Traction (Block 6) is the single most powerful risk-reducer — if present, evaluate how well it’s presented.
  • The deck should make the A→C journey feel inevitable, not just possible.
  • Shorter decks win. Every slide must earn its spot. Two okay slides are weaker than one great slide — always recommend merging when the message overlaps.
  • If a deck exceeds 15 slides, treat length itself as a red flag and explicitly call it out.