Startup scorecard
Agent-based startup coaching skills for Claude Code — every claim labeled FACT/ASSUMPTION/UNKNOWN. Zero-hallucination protocol, 10 skills, one master orchestrator.
npx -y skills add ajstars1/startup-coach --skill startup-scorecardAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Evidence-based startup self-evaluation — 8 dimensions scored 0–5 strictly on documented evidence (not optimism), ending with the two biggest risks and the next evidence to collect. Use as a periodic health check, before demo days or investor conversations, when deciding whether to pivot, or when a founder asks "how are we doing?"
SKILL.md
3.2 KB, as published. Nobody here has run it
Startup Scorecard
A periodic self-evaluation for the opportunity/product being worked on NOW. Its power is the scoring rule: scores follow documented evidence, never confidence. Adapted from evidence-first incubation-program practice.
The 8 dimensions
| Dimension | The question |
|---|---|
| Problem clarity | Is the customer problem sharply framed (mad-lib-tight, in customer language)? |
| Customer evidence | Interview evidence, observation, field notes — documented? |
| Early adopter clarity | Who exactly is the first customer segment? |
| Value proposition strength | Why would someone try this first, over their status quo? |
| Solution feasibility | Can this team build/deliver the solution at this stage? |
| Business viability | Is there a believable path to money / unit economics? |
| Market attractiveness | Is the opportunity worth pursuing and scalable? |
| Experiment / scaling readiness | Are the next risks clear? How will you test them and derisk? |
Scoring bands (evidence-based, strict)
0 = no evidence · 1 = very weak · 2 = weak · 3 = some evidence · 4 = strong evidence · 5 = multiple strong evidence points.
Rules:
- Every score requires an Evidence / notes entry citing something that exists (interview debriefs, logs, transactions, signed LOIs, research with URLs). Nothing written down = 0, regardless of founder certainty.
- Cross-check against the Belief Ledger's evidence grades: a dimension whose support is all E0–E1 (opinions, told stories) caps at 1; E2 (observed behavior/artifacts) caps at 3; scores of 4–5 require E3/E4 (commitments, transactions) from multiple independent sources. One customer/anecdote caps at 2 — n=1 is a direction, not a verdict.
- Honesty is the feature: "Customer evidence: yet to validate — 0" is a useful row, not a shameful one.
The closing prompt (always end with this)
What are the two biggest risks in this opportunity right now, and what evidence do we need next?
Output
Scorecard table (Dimension | Score | Evidence cited | Next evidence to get) → for every score ≤ 3, the cheapest next piece of evidence that would raise it → the two-biggest-risks answer → which companion skills to run next (low problem clarity → /problem-discovery; weak customer evidence → /customer-discovery; shaky market attractiveness → /where-to-play + /secondary-research; fuzzy money story → /business-model-canvas).
Re-run after every ~15–20 interviews or before any major commitment. Score movement over time matters more than any single snapshot.
Ledger discipline (v2): reference Belief Ledger IDs ([A4], [F7]) for every claim; propose new/changed items in a closing "Ledger delta" section rather than restating claims — restated claims drift.
Fill-in templates: TEMPLATES.md in this skill's folder — copy the structures into your project's discovery/ and fill with labels.