MCP Reference Servers
Its Sato Score and verification record. Not a safety, quality or returns grade.
Sato Score
72/100· HighHow open, active & verifiable — not a safety or returns grade.
- Maintenance & liveness
- 35/35
- Code transparency
- 20/20
- Interface verifiability
- 8/10
- Docs & demo
- 0/10
- Listing transparency & provenance
- 7/15
- Independent verification
- 2/10
Deploy path reproduced — install verified in an isolated container (proves installability, not runtime safety).
Next step
Watch MCP Reference Servers
One email if what we observe changes — its repo goes quiet or disappears, or the listing is retired. Nothing else.
Check it from your agent
Add Sato Hub’s MCP server (no key), then ask:
https://satohub.ai/api/mcpRun Sato Hub Preflight on modelcontextprotocol/servers before I use it.Your project?
Prove you control the domain and correct this page — description, links, tags, chains. A claim proves control, not verification: it never changes the Sato Score.
Claim this listing →Anchored
2026-10-15 · 486 listingsThis score was committed to a single Merkle root along with every other score we published that day. The root is signed, published, and archivable by anyone — so if we changed this number tomorrow and said it had always been this, the commitment would contradict us. It proves what we published, not that the score is right.
- Leaf
1c1455eb784f32a39e90e5bac93996bf8c01014e5b46aaa18114b783050e7b8c- Preimage
satohub-score-leaf/v1|2026-10-15|mcp-reference-servers|72|High- Proof
- 9 siblings
- Root
004fa814c17a95aa1bf61786b167b9a311db31782a5b37d9d6d93482c2f9b634- Root signature
- Ed25519, kid b04bd38b
- On-chain
- not attested on-chain — key absent (ATTEST_PRIVATE_KEY is not set)
Verify this yourself
# 1. The proof: record, leaf, leaf_preimage, proof (sibling digests)
curl -s "https://satohub.ai/api/verify/mcp-reference-servers/proof?date=2026-10-15"
# 2. The roots file, fetched separately; take the root for 2026-10-15
curl -s https://satohub.ai/.well-known/sato-score-roots.json
# 3. leaf = SHA-256(leaf_preimage). For each sibling s:
# leaf = SHA-256("satohub-score-node/v1" || min(leaf,s) || max(leaf,s))
# The result must equal the root from step 2.
# 4. Check that root's signature against https://satohub.ai/.well-known/jwks.jsonNo command-line tool is published for score proofs yet; the steps above are the whole check. Compare against the published roots file, not against the root in the proof response — checking a proof only against a root served in the same reply proves nothing. Raw proof: /api/verify/mcp-reference-servers/proof. Method: the house rules.
Evidence
- ✓Deploy path reproduced in an isolated container (docker:node:22-bookworm-slim) on Jul 21, 2026: install passed, entry passed. Verifies installability, not runtime safety or on-chain behavior.
- ▸Source reviewed ↗
Record
- Last checked
- 2026-10-15
- Last code activity (commit or release)
- Oct 11, 2026
- Methodology
- Sato Score methodology →
Verification history
A dated, auditable record of every change to this listing's status, verification, and Sato Score.
- Jul 21, 2026Verified: deploy spec
- Jul 20, 2026Sato Score tier: Medium → HighScore 62 → 70.
- Jul 19, 2026Sato Score tier: High → MediumScore 72 → 62.
- Jul 18, 2026Auto-enriched: docs url, deployment options
- Jul 18, 2026Sato Hub began tracking this listing
- Jul 13, 2026Listed on Sato Hub
Embed this badge
[](https://satohub.ai/verify/mcp-reference-servers?utm_source=badge&utm_medium=embed)Paste into your README. The badge reflects the live Sato Score and links back to this report. Free to use, no attribution required.
Embed the 90-day history
Live preview[](https://satohub.ai/verify/mcp-reference-servers/history?utm_source=badge&utm_medium=embed)Paste into your README. The sparkline is the last 90 days of daily Sato Score readings with the 90-day change; days nobody measured are drawn as gaps, never joined across. It links back to the full dated record. Free to use, no attribution required.
Think this score is wrong?
Check it first: every input behind it is published at /api/export/score-inputs.json and the scorer is a verbatim copy of ours, so the arithmetic can be run on your own machine — reproduce this score. If you maintain MCP Reference Servers and still disagree, say so in public: a dispute is published with your evidence beside our answer, whether or not we agree with it.