zephex
⌘K
Get StartedPricingMCP ToolsDocs
←Back×Sign in
Get StartedPricingMCP ToolsDocs

Documentation

Start with the endpoint and API key flow, then branch into editor setup or tool reference.

GET STARTED

What is MCP?IntroductionQuickstartAPI ReferenceUniversal RequirementsZephex vs Local MCPBest PracticesToken EfficiencyCLAUDE.md TemplateAGENTS.md TemplateMCP EcosystemMarkdown Access

EDITORS

OpenCodeOAuthVS CodeClaude CodeCursorWindsurfJetBrainsKiro CLIOAuthZedOAuthGemini CLIOAuthClineOAuthGooseCodex CLIOAuthFactory AIOAuth

OPERATING SYSTEMS

macOSWindowsLinux

TOOLS

get_project_contextread_codefind_codecheck_packageaudit_packageexplain_architectureZephex_dev_infoscope_taskaudit_headersthinking

SUPPORT

PlansPro & Max GuideUsage & AnalyticsConnection IssuesRate LimitsBillingSecurityFAQChangelog

find_code

Searches the repo for a string or pattern so you can find the right file before you read code.

DESCRIPTION

This is the location tool. It tells you where something lives, not how it works. Use it to find env vars, route names, error messages, feature flags, or scattered call sites before you narrow with read_code.

WHEN TO USE

GOOD FIT

  • When you know the name of what you are looking for but not the file.
  • When you need every occurrence of a string, constant, or route segment.
  • When you want to narrow a repo before calling read_code.

AVOID IT WHEN

  • You already know the exact file and symbol; use read_code.
  • You need a project overview; use get_project_context.
  • You need task-scoped planning; use scope_task.
PARAMETERS
NameTypeRequiredDescription
querystringYesSearch text or regex pattern.
pathstringNoAbsolute path to the repo or subdirectory to search. Omit to search the current project.
DETAILS

PATH PARAMETER

About the path parameter

Use the absolute path to the project root on the machine where the code is available. In Cursor, VS Code, or Windsurf that usually means the workspace root you already have open. On a remote dev box, pass the absolute path on that remote machine. If the path does not exist or the service cannot access it, the tool returns an explicit error instead of guessing. Results are best when the folder is a real project root with .git or package.json. Private GitHub repos are not fetched automatically.

SEARCH STRATEGY

How to get better hits

Start with the most stable string you have: an env var, endpoint segment, error message, or exported symbol name. If the result set is still wide, narrow with a more exact query and then hand the target file to read_code.

EXAMPLE
tools-call.json
{  "jsonrpc": "2.0",  "id": 3,  "method": "tools/call",  "params": {    "name": "find_code",    "arguments": {      "query": "rate_limit_exceeded",      "path": "/Users/name/projects/zephex"    }  }}
OUTPUT
results.txt
dashboard/src/app/docs/api-reference/page.tsx:162 "rate_limit_exceeded"dashboard/src/app/docs/support/billing/page.tsx:74 subsequent tools/call requests return a rate_limit_exceeded errorsrc/middleware/scope-task-rate-limit.ts:41 super("Rate limit exceeded for scope_task")

TOKEN EFFICIENCY

Most find_code calls land around 80 to 250 tokens because they return matches, not file bodies. The manual alternative is opening multiple candidate files, which often turns into 5,000 to 20,000 tokens of unnecessary reading.