Skip to content

Nonsynthetic MCP server

Which MCP servers are Nonsynthetic's own, from the official registry and from asking its own domain, whether they answer, and what they offer. A program checked each of these, nobody copied them from a product page.

read 2026-10-29

← Back to Nonsynthetic

Whose server it is
  • Registry namenet.nonsynthetic/apex-edgeverified in the MCP registry

The registry name sits under nonsynthetic.net. The registry only lets a publisher use that namespace after proving control of the domain through DNS or HTTP.

When we asked itchecked 2026-10-29
  • Stateanswerschecked 2026-10-29
  • Tools443
  • Server calls itselfapex-edge
  • Server version1.0.0
  • Protocol version2025-06-18

On 2026-10-29 the server answered initialize and tools/list without a sign-in. First answer in 325 ms.

Should an agent connect itSkip it
  • Its tools do none of the product's use cases we checked: draft a website from a short questionnaire; deploy and host a site on a subdomain.
  • Checked on 6 of the 7 days it takes to judge reliability.

Once connected

  • Its definitions take about 23040 tokens of context: enable only the tools a task needs.
  • Parameters are untyped or undescribed in 396 of 400 tools: check arguments before calling.

Said by rules from the measurements above. The likely first call is a model's pick from the tool list; read its description before relying on it.

Measured dailydaily checks: 6, the first on 2026-10-29
  • INITE MCP scoreafter a weeka score needs 7 daily checks: 6 so far
  • What an agent can do with it0 of 2 use cases
  • Answeredmeasuring6 of the 7 daily checks it takes to print a share
  • Median first answer180 ms
  • Space in an agent's context≈ 23,040 tokensthe definitions of all its tools (400), which an agent carries on every turn
  • Tool definitionsC · 67 of 100
  • Typed results0 of 400 toolsdeclared outputSchema: results come typed, read once
  • No description an agent can choose by: 1 of 400
  • Parameters without a description: in 396 of 400
  • Not saying whether it only reads or can destroy: 400 of 400

What an agent can do with it

  • cannot · draft a website from a short questionnaire
  • cannot · deploy and host a site on a subdomain
  • not sure · verify site pages against a public hash manifest

Each of the product's use cases, put to Jev against this server's tool list: can an agent do it with these tools?

Worth a look

None of the five warning signs in these definitions.

A program checks the server once a day: initialize and tools/list, no model. Tokens are the definitions' length at about four characters a token. The grade averages six checks of every tool's name, description, schema and annotations; it says how well the tools are described, not whether they work. The score weighs what was measured and nothing else: answering (25%), doing the product's job (25%), definitions (20%), context (10%), use (10%) and how readily a client connects (10%). A part that was not measured is left out, and a hidden instruction caps the score at 40. The state of vendors' MCP servers: method and numbers →

Tools it offers443 tools · read 2026-10-29
  • hash

    $0.0020 — Compute a SHA-256, SHA-384, or SHA-512 digest of UTF-8 text; returns hexadecimal output (default SHA-256).

  • uuid

    $0.0020 — Generate one cryptographically random UUID v4 identifier.

  • base64

    $0.0020 — Encode UTF-8 text as Base64 or decode Base64 to a binary string; input limited to 32768 characters.

  • time

    $0.0020 — utc timestamp

  • json_validate

    $0.0020 — json validity

  • price

    $0.0020 — spot price (Coinbase public)

  • jwt_decode

    $0.0020 — jwt decode-only (no verification)

  • wordcount

    $0.0020 — text stats (chars/words/lines/sentences)

  • slugify

    $0.0020 — url slug

  • semver

    $0.0020 — semver compare {a,b} -> cmp -1|0|1

  • color

    $0.0020 — hex|rgb list -> hex+rgb+hsl

  • cron_next

    $0.0020 — next occurrences of 5-field numeric cron expr (UTC)

  • token

    $0.0020 — crypto-random token (len, alphabet)

  • hmac

    $0.0020 — hmac sign/verify (sha1|sha256|sha384|sha512)

  • calc

    $0.0020 — safe arithmetic eval (+ - * / % ^ parens)

  • url_parse

    $0.0020 — url -> components + query params

  • url_encode

    $0.0020 — percent-encode/decode (component default)

  • ip_info

    $0.0020 — caller network facts (ip, geo, asn, colo)

  • task_class

    $0.0020 — deterministic task→tier classify (NANO→REASONING)

  • date_calc

    $0.0020 — date parse/shift/diff (iso|unix input)

  • csv_parse

    $0.0020 — csv -> rows (rfc4180-lite)

  • case_convert

    $0.0020 — identifier case (camel|pascal|snake|kebab|upper|lower|title)

  • cidr

    $0.0020 — ipv4 cidr math + membership check

  • req_info

    $0.0020 — sanitized request echo (secrets scrubbed)

  • env_validate

    $0.0020 — dotenv parse+schema check (malformed/dupes/empties/missing)

  • regex_test

    $0.0020 — regex pattern+flags+input -> matches+groups

  • junit

    $0.0020 — tests results -> junit xml for CI

  • diff_lines

    $0.0020 — line diff a/b -> ops (LCS-lite, 2000-line bound)

  • pass_strength

    $0.0020 — password entropy heuristic (not a crack proof)

  • markdown_lint

    $0.0020 — markdown static lint (fences/links/heading jumps)

  • board_quote

    $0.0020 — poster-side bounty pricer — class + reward range from live board history

  • token_count

    $0.0020 — estimate LLM token count (cl100k-style heuristic: words, punctuation, unicode runs) — budget a call before paying for it

  • json_diff

    $0.0020 — structural JSON diff — added/removed/changed key paths with values (not a line diff)

  • units

    $0.0020 — unit conversion — length/mass/temp/bytes/time across common units {value, from, to}

  • armory_5g_network

    $0.0100 — armory vendored tool — 5G network (HTTP SKU /api/v1/armory/5g_network)

  • armory_aave_protocol

    $0.0100 — armory vendored tool — Aave protocol (HTTP SKU /api/v1/armory/aave_protocol)

  • armory_ab_testing

    $0.0200 — armory vendored tool — A/B testing (HTTP SKU /api/v1/armory/ab_testing)

  • armory_achievement_system

    $0.0100 — armory vendored tool — Achievement system (HTTP SKU /api/v1/armory/achievement_system)

  • armory_aes_gcm_cipher

    $0.0200 — armory vendored tool — AES-256-GCM encryption (HTTP SKU /api/v1/armory/aes_gcm_cipher)

  • armory_agent_handoff

    $0.0100 — armory vendored tool — structured handoff protocol between agents (HTTP SKU /api/v1/armory/agent_handoff)

  • armory_agent_memory_store

    $0.0100 — armory vendored tool — episodic + semantic memory with recall scoring (HTTP SKU /api/v1/armory/agent_memory_store)

  • armory_agent_trace_otel

    $0.0100 — armory vendored tool — OpenTelemetry tracing for agent runs (HTTP SKU /api/v1/armory/agent_trace_otel)

  • armory_agentic_ai

    $0.0200 — armory vendored tool — Agentic AI (HTTP SKU /api/v1/armory/agentic_ai)

  • armory_ai_alignment

    $0.0200 — armory vendored tool — AI alignment (HTTP SKU /api/v1/armory/ai_alignment)

  • armory_ai_orchestration

    $0.0200 — armory vendored tool — AI orchestration (HTTP SKU /api/v1/armory/ai_orchestration)

  • armory_ai_safety

    $0.0200 — armory vendored tool — AI safety (HTTP SKU /api/v1/armory/ai_safety)

  • armory_android_notifications

    $0.0100 — armory vendored tool — Android notifications (HTTP SKU /api/v1/armory/android_notifications)

  • armory_android_room_db

    $0.0100 — armory vendored tool — Android Room DB (HTTP SKU /api/v1/armory/android_room_db)

  • armory_angular_module

    $0.0100 — armory vendored tool — Angular module with DI (HTTP SKU /api/v1/armory/angular_module)

  • armory_anomaly_detection

    $0.0200 — armory vendored tool — Anomaly detection (HTTP SKU /api/v1/armory/anomaly_detection)

  • armory_ansible_playbooks

    $0.0100 — armory vendored tool — Ansible playbooks (HTTP SKU /api/v1/armory/ansible_playbooks)

  • armory_api_gateway

    $0.0100 — armory vendored tool — API Gateway (HTTP SKU /api/v1/armory/api_gateway)

  • armory_app_store_packager

    $0.0100 — armory vendored tool — App store submission package builder (HTTP SKU /api/v1/armory/app_store_packager)

  • armory_arbitrum_contract

    $0.0100 — armory vendored tool — Arbitrum contract (HTTP SKU /api/v1/armory/arbitrum_contract)

  • armory_artifact_signer

    $0.0100 — armory vendored tool — Build artifact signing and supply-chain attestation (HTTP SKU /api/v1/armory/artifact_signer)

  • armory_artificial_consciousness

    $0.0200 — armory vendored tool — Global-workspace attention model — winner-take-all stream competition over N ticks (HTTP SKU /api/v1/armory/artificial_consciousness)

  • armory_astro_islands

    $0.0100 — armory vendored tool — Astro islands architecture (HTTP SKU /api/v1/armory/astro_islands)

  • armory_automl_pipeline

    $0.0200 — armory vendored tool — AutoML pipeline (HTTP SKU /api/v1/armory/automl_pipeline)

  • armory_autonomous_vehicle

    $0.0200 — armory vendored tool — Waypoint-route planner returning a heading profile (2-D navigation sim) (HTTP SKU /api/v1/armory/autonomous_vehicle)

  • armory_autonomous_vehicles_iot

    $0.0100 — armory vendored tool — Fleet route planner — per-vehicle waypoint tables (local sim) (HTTP SKU /api/v1/armory/autonomous_vehicles_iot)

The first 60 of 443, as the server listed them.

How to connect

Add the server to your MCP client config. Claude Desktop, Cursor and most other clients read this shape:

{
  "mcpServers": {
    "apex-edge": {
      "type": "http",
      "url": "https://api.nonsynthetic.net/mcp"
    }
  }
}

Add it

Claude Code, in a terminal

claude mcp add --transport http nonsynthetic https://api.nonsynthetic.net/mcp

Add to CursorAdd to VS Code

claude.ai, ChatGPT and other clients: add https://api.nonsynthetic.net/mcp as a custom connector in the settings.

OpenAPI spec

387 paths, read 2026-10-29. Open the spec ↗

Each line is something a program found when it went looking: the registry entry, the server's own answer, a spec that parses. A missing line means we did not find it, not that it does not exist.