Changelog

Unreleased

[0.7.0] - 2025-08-17

Added

  • Added an explicit external-tool contract with a configurable SearXNG web-search adapter.
  • Added YAML/schema configuration for tool registration, environment references, bounded requests, and normalized search results.
  • Added prescient search and an annotated web_search configuration example.
  • Added an optional development SearXNG service to docker-compose.yml for runnable web-search examples.
  • Added opt-in search-to-provider context assembly through Prescient.search_and_generate and prescient search --generate.
  • Added POST /v1/search/generate for REST API consumers to opt in to search-result context.
  • Added POST /v1/search for normalized raw external-tool results without AI generation.
  • Added SearchApi as a hosted web-search adapter with engine, location, language, country, API-key, timeout, and result-limit configuration.
  • Added capability groups for ordered tool-adapter fallback on transient connection and rate-limit failures.
  • Documented SearXNG and SearchApi setup and CLI usage side by side.
  • Added explicit --generate opt-in behavior to the web-search example.
  • Added SEARXNG_URL environment defaults for automatic web_search registration.
  • Expanded prescient config example with documented SearXNG settings, environment references, custom tool names, and direct-versus-generated usage.
  • Organized CLI help into global and search-specific option sections.

[0.6.0] - 2025-08-15

Added

  • Added a dependency-free Rack-compatible REST API with generation, embeddings, bounded batch embeddings, provider/model discovery, capabilities, health, liveness, readiness, version, request IDs, authentication hooks, and JSON error envelopes.
  • Added a small Rack example that lists the REST API endpoints and delegates requests to Prescient::API.
  • Added an optional rack_example bundle group with Rack, Rackup, and Puma for running the example application without adding web-server dependencies to the library.
  • Added a non-root, healthchecked Docker image and Compose example for the REST API, with optional GHCR publication on version tags.
  • Documented mounting Prescient::API in Rails routes with its endpoint catalog and authentication example.
  • Made the CLI and REST API optional lazy-loaded entry points so library users requiring only prescient do not load either interface eagerly.

[0.5.0] - 2025-08-14

Added

  • Added a versioned YAML configuration loader with environment-variable references, configuration validation, precedence rules, and a packaged JSON Schema.
  • Added YAML-configurable prompt templates and CLI prompt overrides, including support for loading multiline templates from a file.
  • Added prescient config example for generating an annotated schema-backed YAML configuration starter.
  • Added Google Gemini provider support for text generation, embeddings, health checks, and model listing.
  • Added Gemini environment-variable defaults and YAML configuration support.
  • Added Mistral provider support for text generation, embeddings, health checks, and model listing.
  • Added Mistral environment-variable defaults and YAML configuration support.
  • Added DeepSeek provider support for text generation, health checks, and model listing.
  • Documented DeepSeek's unsupported embedding capability explicitly.
  • Added xAI provider support for text generation, health checks, and model listing.
  • Documented xAI's unsupported embedding capability explicitly.

Changed

  • Raised YARD documentation coverage enforcement from 99% to 100%.

[0.4.0] - 2025-08-14

Added

  • Added the prescient CLI with provider listing, health checks, configuration validation, generation, embeddings, JSON output, stdin input, and exit-status handling.

Changed

  • Added CLI model overrides to provider generation and embedding operations.
  • Added secure CLI credential sourcing with --api-key-env, alongside direct --api-key support for ephemeral automation.
  • Documented all CLI automation overrides, including provider selection, generic and task-specific models, API keys, environment-backed credentials, and JSON output.

[0.3.0] - 2025-08-14

Added

  • Added a development quality harness covering tests, RuboCop, YARD, and RBS tasks.
  • Added 99% minimum line and branch coverage requirements with expanded provider and context coverage tests.
  • Added an actionlint Rake task for GitHub Actions validation.
  • Added Dependabot configuration for Bundler, Docker, and GitHub Actions updates.
  • Added an opt-in Prescient::Pgvector::Store boundary for validated embedding storage and similarity search.
  • Added configurable Ollama embedding dimensions while preserving strict validation when configured.
  • Added a default example-syntax quality gate that validates Ruby examples without contacting providers.

Changed

  • Updated Ollama embeddings to use /api/embed with strict vector-dimension validation.
  • Restricted fallback to transient/provider-service failures and added the public ProviderError exception for provider-side service errors.
  • Reused registered provider instances and removed redundant health checks during fallback discovery.
  • Updated default chat model names and README examples for current Ollama, Anthropic, OpenAI, and Hugging Face model selections.
  • Updated Hugging Face inference to use the current router feature-extraction and OpenAI-compatible chat-completion APIs.
  • Standardized provider reachability in health results and removed embedding padding/truncation across OpenAI and Hugging Face.
  • Updated Anthropic model listing and health checks to use its /v1/models catalog endpoint.
  • Retained OpenAI Chat Completions for the current normalized public response contract; Responses API migration remains a separately scoped compatibility change.
  • Raised YARD API documentation coverage enforcement to 99% or higher and modernized gem development metadata.
  • Compacted GitHub Actions into focused CI, Pages, release, and security workflows.
  • Added credentialed, provider-selected live smoke tests that remain skipped by default.
  • Restored the RBS/Steep development tasks with a committed Steepfile and curated core API signatures.
  • Expanded Steep coverage to the base abstraction and all provider adapters, including their public operations and HTTP boundaries.
  • Added configurable provider-info sensitive-key sanitization and configurable generic context-field exclusions.
  • Audited README and documentation examples against the current codebase.
  • Removed client method_missing delegation so provider-specific behavior is not exposed through the public client.
  • Prevented default provider registration when required credentials are absent from the environment.
  • Sanitized provider HTTP errors so raw response bodies are not exposed in exception messages.
  • Removed fallback health probes before provider operations to avoid duplicate network requests and race conditions.
  • Clarified health semantics: reachable reports transport availability and ready reports configured-model readiness.
  • Added shared provider health-contract coverage for status, reachability, readiness, and provider identity.

Removed

  • Removed the obsolete PDF changelog and superseded workflow definitions.

[0.2.0] - 2025-08-05

Added

  • Added new featire: Providers fallbacks mechanism

[0.1.0] - 2025-08-05

Added

  • Initial release of Prescient gem
  • Support for four AI providers:
    • Ollama: Local AI provider with embedding and text generation
    • Anthropic: Claude models for text generation
    • OpenAI: GPT models and embeddings
    • HuggingFace: Open-source models and embeddings
  • Unified client interface for all providers
  • Comprehensive error handling with provider-specific exceptions:
    • ConnectionError for network issues
    • AuthenticationError for API key problems
    • RateLimitError for rate limiting
    • ModelNotAvailableError for missing models
    • InvalidResponseError for malformed responses
  • Automatic retry logic with configurable attempts and delays
  • Health monitoring capabilities for all providers
  • Environment variable configuration support
  • Programmatic configuration system
  • Context-aware generation support with context items
  • Text preprocessing and embedding normalization
  • Provider availability checking
  • Model listing capabilities (where supported)
  • Comprehensive test suite with RSpec
  • Documentation and usage examples

Provider-Specific Features

  • Ollama: Model management (pull, list), local deployment
  • Anthropic: Latest Claude 3 models (Haiku, Sonnet, Opus)
  • OpenAI: Multiple embedding dimensions, latest GPT models
  • HuggingFace: Open-source model support, research-friendly API

Development

  • RSpec test suite with WebMock and VCR
  • RuboCop code style enforcement
  • SimpleCov test coverage reporting
  • Comprehensive documentation
  • Example usage scripts
  • Rake tasks for testing and linting

[0.0.0] - 2025-08-05

Added

  • Project initialization
  • Basic gem structure
  • Core interfaces defined