Add comprehensive documentation and guides for Cycles

[amavashev]

Summary

This PR adds extensive documentation and guides for the Cycles budget governance system, including quickstart tutorials, operational guides, integration patterns, and incident postmortems. The documentation covers deployment, configuration, integration with various LLM providers, and production best practices.

Key Changes

New Quickstart Guides:

• quickstart/what-is-cycles.md - Introduction to Cycles concepts and the reserve-commit lifecycle
• quickstart/end-to-end-tutorial.md - Complete 10-minute tutorial from zero to a working budget-guarded LLM call, including Docker Compose setup, tenant creation, budget funding, and sample applications in Python and TypeScript

Operational & Production Guides:

• how-to/production-operations-guide.md - Redis configuration, persistence, HA setup, Cycles Server configuration, network architecture, and monitoring
• how-to/monitoring-and-alerting.md - Key metrics (budget utilization, reservation lifecycle, server health), alerting thresholds, and observability patterns
• how-to/security-hardening.md - Network isolation, Redis security, API key management, least-privilege permissions, and audit logging

Integration Guides:

• how-to/integrating-cycles-with-express.md - Middleware and inline patterns for Express.js applications
• how-to/integrating-cycles-with-vercel-ai-sdk.md - Integration with Vercel AI SDK for Next.js streaming applications
• how-to/integrating-cycles-with-aws-bedrock.md - Non-streaming and streaming patterns for AWS Bedrock
• how-to/integrating-cycles-with-google-gemini.md - Integration with Google Gemini API

Practical How-To Guides:

• how-to/choosing-the-right-integration-pattern.md - Decision tree and comparison of decorator, streaming adapter, middleware, and programmatic client patterns
• how-to/cost-estimation-cheat-sheet.md - USD_MICROCENTS unit explanation, provider pricing reference tables, and quick estimation formulas
• how-to/common-budget-patterns.md - Recipes for per-user daily budgets, per-conversation budgets, model-tier budgets, and team-level rollup budgets
• how-to/adding-cycles-to-an-existing-application.md - Incremental adoption path with shadow mode, wrapping individual calls, and expanding coverage
• how-to/troubleshooting-and-faq.md - Common issues (budget exceeded, TTL expiry, idempotency mismatches) with solutions

Incident Postmortems:

• incidents/scope-misconfiguration-and-budget-leaks.md - Failure mode from inconsistent scope configuration
• incidents/retry-storms-and-idempotency-failures.md - Unbounded retry loops causing overspend
• incidents/concurrent-agent-overspend.md - Race conditions with concurrent agents sharing budgets

Reference:

• changelog.md - Release history and protocol specification summary for v0.1.23

Documentation Site Updates:

• Updated .vitepress/config.ts to reorganize navigation and add new guide sections
• Updated index.md homepage with improved tagline and call-to-action

Notable Implementation Details

• Documentation uses consistent code examples across Python, TypeScript, and Java SDKs
• Pricing tables include current rates for OpenAI, Anthropic, and Google models with microcent conversions
• Guides include complete, runnable examples (Docker Compose configs, curl commands, application code)
• Troubleshooting guide covers the most common integration issues with step-by-step fixes
• Incident postmortems provide real-world failure scenarios and prevention strategies
• Navigation structure organizes content into Quickstart, How-To, and Reference sections

https://claude.ai/code/session_016sFHtxX3RSCvHyu7eV4nuH