HISS Score
userpaper modeThe transparency and structure heuristic behind every basket's score.
What the score is
Every basket gets a HISS Score from 0 to 100: the mean of six component scores (seven when an MCP plan is attached), each also 0–100. It is a transparency and structure heuristic, never a prediction — it measures how well a thesis is expressed and disclosed, not whether it will make money.
Components
| Component | Field | What it measures |
|---|---|---|
| Thesis clarity | thesisClarity | Length and structure of the written thesis — a fleshed-out argument scores higher than a one-liner. |
| Diversification | diversification | Effective number of holdings via the Herfindahl index; eight or more effective positions maxes it out. |
| Liquidity readiness | liquidityReadiness | Share of holdings bound to verified canonical addresses (placeholder addresses score zero). |
| Oracle health | oracleHealth | Fraction of holdings whose price feed currently reports OK. |
| Meme strength | memeStrength | Tagline, emoji, hashtags. Deliberately shallow — shareability, not substance. |
| Risk disclosure | riskDisclosure | Presence and completeness of risk notes plus the mandatory not-investment-advice and canonical-only flags. |
| MCP safety | mcpSafety (optional) | Only for baskets with an MCP plan: no-autotrade default, required user review, and a review checklist. |
Ranges
- 0–39 — structurally weak: thin thesis, missing disclosures, or unpriceable holdings.
- 40–69 — solid: a real thesis with honest risk notes and mostly healthy plumbing.
- 70–100 — well-built: clear argument, meaningful diversification, verified assets, complete disclosure.
Venom: risk intensity
Venom — how concentrated and aggressive a basket is — is not a score component. It is described in the risk summary generated alongside every score, which names the largest position, the concentration cap, and drawdown-relevant structure. A high-Venom basket can score high if it is honest about what it is; the score rewards disclosure, not caution.
Simulated performance (paper PnL and max drawdown) is appended to the score's explanation string for context, clearly labeled paper-only — it never feeds the component math.