Skip to content

CHANGELOG.md

Latest commit

 

History

History
203 lines (165 loc) · 7.8 KB

CHANGELOG.md

File metadata and controls

203 lines (165 loc) · 7.8 KB
Raw

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project follows Semantic Versioning.

Unreleased

Added

1.9.8 - 2026-01-05

Fixed

  • CRITICAL: Fixed false positives for .ai TLD (domains reported as "available" when actually registered)
  • Added WHOIS cross-verification for unreliable RDAP TLDs (.ai, .io, .sh, .ac)
  • Implemented native TCP WHOIS lookup for accurate .ai/.io/.sh/.ac domain status

Changed

  • RDAP 404 responses for .ai/.io/.sh/.ac now trigger WHOIS verification
  • WHOIS module now uses native TCP (port 43) for TLDs with unreliable web APIs

1.9.6 - 2026-01-05

Added

  • Hybrid Redis + in-memory cache for horizontal scaling
  • Circuit breaker with automatic Redis fallback to in-memory
  • Connection health monitoring with exponential backoff reconnection

Changed

  • Cache layer now uses Redis as primary with seamless in-memory fallback

1.9.5 - 2026-01-04

Added

  • ai_health tool for monitoring AI inference service status
  • Circuit breaker state visibility (closed/open/half-open)
  • Adaptive concurrency metrics exposure

1.9.4 - 2026-01-04

Added

  • Adaptive batch sizing for AI inference based on response times
  • Dynamic concurrency adjustment (1-4 parallel requests)
  • Batch size optimization (1-5) based on latency percentiles

1.9.3 - 2026-01-04

Added

  • Circuit breaker pattern for AI inference resilience
  • Parallel AI suggestion generation with Promise.allSettled
  • Automatic fallback to semantic engine on AI failures

Changed

  • AI suggestion flow now uses parallel requests with circuit breaker protection

1.9.2 - 2026-01-03

Added

  • /metrics endpoint with Prometheus-compatible format
  • Cache hit/miss statistics per cache type
  • Request latency histograms
  • AI inference success/failure counters

1.9.1 - 2026-01-02

Added

  • HTTP transport rate limiting (100 requests/minute per IP)
  • Rate limit bypass for /health endpoint
  • Connection tracking and cleanup on disconnect

1.9.0 - 2026-01-01

Added

  • HTTP/SSE transport mode (--http flag or MCP_TRANSPORT=http)
  • OpenAPI 3.1 specification auto-generation from Zod schemas
  • REST API endpoints (/api/tools/*) for ChatGPT Actions compatibility
  • Session management with UUID tracking
  • CORS support with configurable origins

Changed

  • Server now supports dual transport: stdio (default) and HTTP

1.8.1 - 2025-12-31

Fixed

  • Sync SERVER_VERSION constant with package.json version

1.8.0 - 2025-12-31

Added

  • Zero-config AI-powered suggestions via public Qwen 7B-DPO inference
  • suggest_domains_smart tool with fine-tuned domain generation
  • OpenRouter fallback for AI inference redundancy
  • Security hardening for inference endpoints

Changed

  • AI suggestions now work out of the box without API keys

1.6.2 - 2025-12-31

Fixed

  • CI/CD release workflow for proper npm publishing with provenance

1.5.1 - 2025-12-31

Fixed

  • RDAP false positives corrected using backend availability signal

Removed

  • GoDaddy direct API integration (replaced with backend)

1.5.0 - 2025-12-31

Added

  • Federated negative cache system for pre-filtering taken domains
  • expiring_domains tool for monitoring domains approaching expiration
  • Cache backend deployment scripts for VPS

1.4.0 - 2025-12-31

Added

  • Security improvements for API key handling
  • Performance optimizations for bulk searches

1.3.2 - 2025-12-31

Fixed

  • WHOIS API error response handling

1.3.1 - 2025-12-31

Fixed

  • SERVER_VERSION sync with package.json

1.3.0 - 2025-12-31

Added

  • analyze_project tool for scanning local/GitHub projects and suggesting matching domains
  • hunt_domains tool for domain investment opportunities (Sedo auctions, pattern generation, investment scoring)
  • Project manifest parsing (package.json, pyproject.toml, Cargo.toml, go.mod)
  • README keyword extraction for better suggestions

1.2.32 - 2025-12-30

Fixed

  • who.is parser fix for production

1.2.31 - 2025-12-30

Fixed

  • who.is parser updated for Next.js HTML format changes

1.2.30 - 2025-12-30

Fixed

  • WHOIS fail-safe on ambiguous API responses

1.2.29 - 2025-12-30

Added

  • llama.cpp integration for local Qwen AI domain suggestions

1.2.28 - 2025-12-30

Changed

  • Refined smart suggestion filtering for crypto/tech prompts to avoid fragment outputs.
  • Prefer Porkbun price check links when pricing backend is enabled and registrar is unknown.

1.2.27 - 2025-12-30

Added

  • Nameserver-based aftermarket hints (Sedo/Dan/Afternic) with cache + timeout controls.
  • RDAP bootstrap caching with IANA fallback reuse.
  • Compact table output with pricing labels and link consolidation.
  • Tests for Sedo parsing, RDAP bootstrap cache, and table formatting.

1.2.26 - 2025-12-30

Added

  • Release Drafter automation for release notes (labels -> changelog).

1.2.25 - 2025-12-30

Added

  • GitHub Actions release workflow that publishes with provenance and creates GitHub Releases.

1.2.24 - 2025-12-30

Added

  • Release workflow documentation and local publish scripts (canary/latest).

1.2.23 - 2025-12-30

Added

  • Sedo public feed lookup for aftermarket auction hints (configurable TTL + feed URL).

1.2.22 - 2025-12-30

Changed

  • Removed Dynadot backend usage due to ToS restrictions.