SqueezeRank MSO
Score, audit, optimize and benchmark MCP listings before agents have to guess.
/v1/mso/card and /v1/mso/report/<id> routes are not counted as Marketplace tools.SqueezeRank MSO is ScriptMasterLabs’ MCP Store Optimizer: a service that scores any MCP server listing across six discoverability axes and returns per-tool grades A+ to F, a weighted final score, and a prioritized fix list. Six tools, five paid per-call in USDC on Base via x402 ($0.002–$0.010), one free.
Before launch, SqueezeRank MSO ran its own six-axis scorer against its own listing. Result: A (4.55/5.0) — computed by the same deterministic engine it sells, not asserted in marketing copy. The scorer is deterministic: same manifest in, same score out, every time. The product optimized itself first.
It then scored SqueezeOS — the live MCP server this product runs on. Result: B- (2.94/5.0) across 13 real tools. Top fixes: no when-to-use guidance on any tool, server description over 20 words, most parameters undescribed. That gap is the product: the difference between what agents see and what they should see.
A full optimization cycle costs $0.023 USDC: score_listing $0.003, optimize_description $0.002, audit_tools $0.005, benchmark_registry $0.003, generate_fix_plan $0.010 — with a receipt for every step. GET /v1/mso/pricing is free. All five paid calls settle in USDC on Base mainnet (eip155:8453) via x402.
Purpose Clarity (25%), Usage Guidelines (20%), Behavioral Transparency (20%), Parameter Semantics (15%), Conciseness & Structure (10%), Contextual Completeness (10%). Server score = 0.70 × (0.60 × avg tool score + 0.40 × min tool score) + 0.30 × coherence: one badly described tool drags the whole server down.
What it does
Scores MCP server listings across six discoverability axes, audits individual tool descriptions, drafts improved server descriptions, checks public registry presence, and builds a prioritized fix plan. All suggested rewrites are drafts for human review.
Machine-readable OpenAPI 3.0.3 specification · Live pricing JSON · Live server card
Five paid tools
POST /v1/mso/score— $0.003 USDC — six-axis score, per-tool grades and prioritized fixes.POST /v1/mso/optimize— $0.002 — three draft server descriptions ranked best-first.POST /v1/mso/audit— $0.005 — per-tool audit plus draft rewrites.POST /v1/mso/benchmark— $0.003 — public registry probes reported as listed, missing or unverifiable.POST /v1/mso/fix-plan— $0.010 — prioritized week-by-week optimization roadmap.
One call to each of the five paid tools totals $0.023 USDC. GET /v1/mso/pricing is free. No subscription.
The six scoring axes
| Axis | Weight | Measures |
|---|---|---|
| Purpose Clarity | 25% | Capability-first wording and keyword clarity |
| Usage Guidelines | 20% | When-to-use guidance |
| Behavioral Transparency | 20% | Read-only/destructive/idempotent/cost signals |
| Parameter Semantics | 15% | Input descriptions |
| Conciseness & Structure | 10% | Dense, consistent listing copy |
| Contextual Completeness | 10% | Returns, errors, examples and links |
Evidence and limits
Internal deterministic self-test: A (4.55/5.0), computed 2026-10-04 by the product's own scorer. This is internal product evidence, not third-party validation.
SqueezeOS audit: B- (2.94/5.0) against the live 13-tool manifest on 2026-10-04.
Not a rank tracker. Not a registry. Not a promise of traffic. It does not invent ecosystem percentiles. Registry checks return observed listed/missing/unverifiable states; unreachable sources remain unverifiable.
Status
Live. The Circle submission should be read as six Marketplace tool endpoints: five paid calls and the free pricing endpoint. The free card and report-reader routes are auxiliary discovery/evidence surfaces. Paid calls use x402 in USDC on Base mainnet (eip155:8453).
Related products
SqueezeOS MCP · SML Commit · SML Notary · Decision-gated payments · URL-to-Markdown API
SCRIPTMASTERLABS · READ · EXPLORE · VERIFY · EARN