| name | vc-lens |
| description | Review a captured business idea from a venture-capital lens — market size, defensibility/moat, team-market fit, capital efficiency, exit pathways, traction milestones, and investability. Produces a structured analysis with a clear verdict. Use when the user wants to know how a VC would react to the idea. |
VC Lens
Apply a VC-style review to a single idea. Output is a markdown analysis document, not a recommendation.
Inputs
- Idea slug (must already exist in
<workspace_path>/ideas/<slug>/)
- The idea's
idea.md and any versions/*.md variants
- User context from
$CLAUDE_USER_DATA/business-idea-eval/user-context.md
Analysis dimensions
Address each in 1–3 paragraphs:
- Problem & customer — pain severity, frequency, who actually pays, willingness-to-pay signal
- Market — TAM/SAM/SOM rough sizing (defer detail to
tam-analysis), category maturity
- Wedge & differentiation — what's the initial wedge, why now, what's the unfair advantage
- Moat — network effects, data, regulatory, switching cost, brand, scale economies — or honest absence
- Business model — unit economics intuition, pricing power, GTM motion (PLG / sales-led / hybrid)
- Capital intensity — bootstrap-viable, seed-friendly, or capital-hungry; rough burn-to-milestone estimate
- Team-market fit — is the founder (the user, per user-context.md) credibly positioned to win this
- Exit landscape — strategic acquirers, public-market path, or lifestyle business — set expectations honestly
- Risks — top three, ranked, with severity
- Verdict — investable as-is / investable with changes / fund-of-friends or bootstrap only / pass
Output
Write to <workspace_path>/ideas/<slug>/analysis/vc-lens.md with frontmatter:
---
lens: vc-lens
generated: YYYY-MM-DD
verdict: <one of the four verdicts>
---
Notes
- Calibrate to the user's actual situation per
user-context.md. Don't tell a bootstrapper "raise a Series A" if they explicitly want to stay self-funded.
- Be direct. Empty validation is worse than honest pushback.
- Cross-reference
reference/scorecard.md (ICEC) if the user wants quantitative scoring layered on top.
