zephex
CLIGet StartedPricingMCP ToolsCommunityGuidesDocs
←BackSign in
CLIGet StartedPricingMCP ToolsCommunityGuidesDocs
Get started freeSign in
DocsAPIToolsEditorsChangelogHelp

GET STARTED

WelcomeQuickstartSetup videoMCP Q&A (learn)BlogWhat is MCP?Who is Zephex for?Plans & PricingZ-GASAB benchmarkBenchmark chart (live)Changelog

CLI

Install & upgradeStartAll install methodsnpx zephex setupCLI without an agentFirst run (init)Login, logout, keysCommand referenceTerminal Mode 2Slash commandsCLI in Docker

INSTALLATION

Connect MCPAll editors (one command)Crush, Hermes, ChatGPT, KiloVS Code Marketplace extensionOAuth & HTTP setupSetup walkthroughHTTP vs stdioWeb Terminal (dashboard)Web Terminal toolsTerminal tools (complete)Test PulseTest Pulse commandsdeep (Project Dossier)Project MemorySupply PulseSupply Pulse commandsinstall (GitHub / npm files)install file commandsCommand CompassHTTP API

API & KEYS

API Key ManagementKey Naming & FormatAuthenticationKey Dashboard

CONFIGURATION

Universal RequirementsSupported EditorsHow It WorksArchitectureCLAUDE.md TemplateAGENTS.md Template

EDITORS28 guides

Supported EditorsVS CodeVS Code extension (Marketplace)Claude CodeCursorWindsurfJetBrains

PLATFORM

macOSWindowsLinux

TOOLS10 tools

Capabilities OverviewTools OverviewTool FilteringTool Workflowsget_project_contextread_codefind_codecheck_packageexplain_architectureZephex_dev_infocheck_testaudit_headerskeep_thinkingproject_memory

GUIDES

Best PracticesToken EfficiencyUse CasesZephex vs Local MCPZephex vs Context7Zephex vs GitHub MCPZephex vs SmitheryMCP EcosystemMarkdown Access

SUPPORT

Help CenterMCP troubleshootingTeam rolloutFAQConnection IssuesRate LimitsDowntime & ErrorsBillingTier GuidePro & Max guideUsage LimitsUsage Analytics

LEGAL

System StatusTerms (summary)Privacy (summary)Data UseSecurityAuthenticationSecurityData HandlingPrivacy PolicyTerms of Service

Quick Links

API Reference

Complete API documentation

Troubleshooting

Common issues and solutions

Community

Join our Discord community

Plugins

Editor and CLI integrations

Pricing

Free, Pro, and Max plans

Enter
Zephex_devzephex-devzephexzephexhello@zephex.dev
© 2026 Zephex. All systems operational.
Docs/Reviewer

Reviewer test instructions

For Anthropic Connectors Directory and OpenAI Apps SDK reviewers evaluating Zephex.

Quick facts

MCP endpointhttps://zephex.dev/mcp
Auth typeOAuth 2.0 (Anthropic-held credentials for Directory)
Authorization serverhttps://zephex.dev
Tool count10 tools, all read-only
Protocol version2025-11-25
Egress allowlistNot needed — zephex.dev is fully public

Test account

Test credentials are shared securely via the submission form attachment. The reviewer email is anthropicreviewer-zephex@inbox.testmail.app (Pro tier, active subscription, no Stripe charges). Password and pre-provisioned API key are in the secure attachment.

Step-by-step OAuth test

  1. 1. In Claude.ai (web/Desktop/mobile), open Settings → Connectors → Add custom connector.
  2. 2. Enter URL: https://zephex.dev/mcp
  3. 3. Click Connect. Browser opens to https://zephex.dev/oauth/authorize → /cli/auth?session=…&flow=mcp.
  4. 4. Sign in with the reviewer email and shared password (Google, GitHub, or email).
  5. 5. Approve access when prompted.
  6. 6. You're returned to Claude with the connector connected.

Tool tests (all should succeed)

audit_headers

Prompt: "Use audit_headers on https://example.com"
Expected: A+ to F security grade, list of headers, fix snippets.

check_package

Prompt: "Use check_package to verify that 'react' exists on npm"
Expected: { exists: true, deprecated: false, latest_version: "19.x.x" }

project_memory

Prompt: "Use project_memory to remember this project's conventions"
Expected: Persistent recall of project decisions, gotchas, and conventions.

Zephex_dev_info

Prompt: "Use Zephex_dev_info to search for OAuth refresh tokens"
Expected: Either a knowledge base article result or a graceful "no entry found" message.

keep_thinking

Prompt: "Use keep_thinking with thought: 'Test reasoning step', thoughtNumber 1, totalThoughts 1, nextThoughtNeeded false, confidence 0.9, thoughtType observation, goalAnchor 'verify tool works'"
Expected: { thoughtConfirmed: "[1/1] Test reasoning step", shouldContinue: false }

Project tools (get_project_context, check_test, read_code, find_code, explain_architecture)

These read project files. From hosted Claude, pass inline_files (a JSON object mapping relative paths to file contents) or a GitHub URL like github:vercel/next.js.
Example prompt: "Use get_project_context with this package.json: {"name":"test","dependencies":{"express":"^4.18.0"}}"
Expected: detected stack ("JavaScript / Express / Node.js"), dev_command, entry_point.

Spec compliance

  • ✅ MCP 2025-11-25 protocol
  • ✅ RFC 9728 Protected Resource Metadata at /.well-known/oauth-protected-resource
  • ✅ RFC 8414 Authorization Server Metadata at /.well-known/oauth-authorization-server
  • ✅ OIDC Discovery 1.0 at /.well-known/openid-configuration
  • ✅ PKCE S256 supported
  • ✅ Token endpoint accepts application/x-www-form-urlencoded
  • ✅ Refresh token rotation enabled
  • ✅ WWW-Authenticate header on 401 with realm, resource_metadata, scope
  • ✅ All 10 tools have title + readOnlyHint: true + destructiveHint: false
  • ✅ Origin header validation enforced (allowlist: claude.ai, claude.com, chatgpt.com, platform.openai.com, zephex.dev)
  • ✅ HTTPS / TLS 1.3 only (HSTS preloaded)
Questions during review? support@zephex.dev