README
01Agent Claude Code
Agents are invoked via Claude Code's Task tool with the appropriate agent file.
Agent Claude Code
Agents are invoked via Claude Code's Task tool with the appropriate agent file.
Agent Claude Code
Meta-agent for optimizing Claude Code configuration and MCP server usage.
Agent Claude Code
Multi-perspective code review. Runs type checking, linting, MCP compliance, Google API best practices, security scan, and test coverage checks in a single pass. Catches 95% of issues before CI. Use before commits for fast feedback (<2min).
Agent Claude Code
Execution path tracer for ServalSheets. Traces the 4-layer pipeline (STDIO→server.ts→tool-handlers.ts→handlers/.ts→google-api.ts) to pinpoint where failures originate. Use when tests fail unexpectedly, behavior doesn't match the schema, or to understand request flow through the system.
Agent Claude Code
Coordinates ServalSheets development by orchestrating the right specialist agents in the right order. Give it a task and it handles research → implementation → validation → review. Examples: 'implement ISSUE-047', 'fix the failing test in composite.ts', 'review everything before committing', 'what broke the tests?'…
Agent Claude Code
Use this agent when you need expert-level guidance on Google APIs (Sheets, Drive, Apps Script, BigQuery) within the ServalSheets codebase. Trigger this agent when: implementing new Google API integrations, optimizing existing API usage patterns, auditing handler code against official Google documentation, debugging…
Agent Claude Code
Google Sheets API v4 expert with real-time documentation access. Validates API usage patterns, quota optimization, best practices, and error handling. Always checks latest Google documentation for accuracy. Use when implementing new actions, debugging API issues, or optimizing quota usage.
Agent Claude Code
Google Apps Script API expert for custom function and automation patterns.
Agent Claude Code
Google BigQuery API expert for Sheets-BigQuery integration patterns.
Agent Claude Code
Google Drive API v3 expert with real-time documentation access.
Agent Claude Code
MCP protocol compliance expert for ServalSheets. Validates protocol adherence, checks transport implementations, verifies SDK compatibility, and ensures spec compliance with MCP 2025-11-25. Use when implementing new tools, modifying server handlers, or debugging protocol issues.
Agent Claude Code
Use this agent when you need expert guidance on the Model Context Protocol (MCP) specification, compliance validation, protocol implementation questions, or when reviewing code that interacts with MCP internals. Examples include:\n\n \nContext: Developer is implementing a new MCP transport layer in…
Agent Claude Code
Performance optimization specialist for ServalSheets. Profiles code, identifies bottlenecks, optimizes quota usage, implements caching strategies, and validates performance regressions. Always uses benchmarks and real metrics. Use when optimizing handlers, debugging slow operations, or reducing API costs.
Agent Claude Code
Security review specialist for ServalSheets. Audits for OWASP issues, OAuth/credential handling, SQL injection in BigQuery queries, input sanitization, API key exposure, and authorization gaps. Use before releases or when adding auth/data-handling features.
Agent Claude Code
Elite MCP QA agent. Tests all 25 tools + 410 actions end-to-end. Executes live API test suite, validates MCP compliance across all tools, catches integration gaps. Examples: 'run comprehensive test suite', 'execute live API tests', 'validate all tools', 'check MCP compliance for all 25 tools'.
Agent Claude Code
TDD-based feature implementation for ServalSheets. Use for adding new actions, fixing bugs, implementing features, or modifying handlers/schemas. Always writes tests first, follows existing patterns, and validates with gates. Examples: Add bulkdelete action to sheetsdata; Fix composite streaming test failures…
Agent Claude Code
Automated MCP protocol tester for ServalSheets. Runs all 410 actions via actual JSON-RPC protocol (not direct handler calls), captures results, analyzes failures, and suggests fixes. Use instead of manual MCP Inspector UI. Examples: 'run protocol smoke test', 'test sheetsdata actions via MCP', 'analyze what failed in…
Agent Claude Code
Fast codebase research for ServalSheets using read-only operations. Use for pattern discovery, finding code examples, analyzing handlers/schemas, counting occurrences, or extracting architectural patterns. Examples: Find all error handling patterns in handlers; Analyze action naming conventions across 25 tools; Find…
Agent Claude Code
Fast automated validation using ServalSheets gate pipeline (G0-G5 and A1-A15). Use for pre-commit checks, schema change validation, regression detection, or phase completion verification. Always uses Haiku for cost efficiency. Examples: Run G0 baseline validation; Validate schema changes; Full gate pipeline for phase…
Agent Claude Code
Comprehensive testing specialist for ServalSheets. Designs test strategies, writes property-based tests, implements mutation testing, and ensures 100% critical path coverage. Uses TDD/BDD patterns and fast test execution. Use when adding features, fixing bugs, or improving test coverage.
Skill Claude CodeCodex
Trace a ServalSheets failure through the 4-layer pipeline to find the exact origin. Use when tests fail unexpectedly, behavior doesn't match schema, or a specific error needs root cause analysis.
Skill Claude CodeCodex
Enterprise-grade Google Sheets MCP server (ServalSheets) with tool categories and specialized actions. Implements UASEV+R protocol for intelligent spreadsheet operations with transaction support, AI analysis, conversational context, and MCP 2025-11-25 compliance.
Skill Claude CodeCodex
Implement a ServalSheets feature or fix using the full research → implementation → validation → review pipeline.
Skill Claude CodeCodex
Run a comprehensive multi-perspective code review on staged or recent changes. Covers TypeScript types, MCP compliance, Google API best practices, security, and test coverage.