TODO.md

Go-LLMs Project TODOs - v0.3.x Roadmap

v0.3.1 Release ✅ COMPLETED (January 10, 2025)

v0.3.2 Release ✅ COMPLETED (January 11, 2025)

v0.3.2 Documentation simplification ✅ COMPLETED (January 11, 2025)

v0.3.3 Additional Providers in llm library

0.3.3.1: Ollama provider ✅ COMPLETED (January 11, 2025)

0.3.3.2: OpenRouter provider ✅ COMPLETED (January 11, 2025)

0.3.3.3: Google Vertex AI provider ✅ COMPLETED (January 11, 2025)

0.3.3.4: Mistral AI provider (https://mistral.ai/) 🔄 DEFERRED TO v0.4.x

• Research Mistral AI API and update this todo.md list
  • Investigate API authentication method
  • Check API format (custom or OpenAI-compatible)
  • Document available models (Mistral 7B, Mixtral, etc.)
  • Identify streaming support and special features
  • Check for model listing/discovery endpoints
• Add dedicated provider implementation
  • Create pkg/llm/provider/mistral.go
  • Add Mistral-specific options if needed
  • Write unit tests in pkg/llm/provider/mistral_test.go
• Add model discovery/listing support (if available)
  • Implement fetcher in pkg/util/llmutil/modelinfo/fetchers/mistral_fetcher.go
  • Add tests for the fetcher
  • Integrate with modelinfo service
• Create dedicated example in cmd/examples/provider-mistral/
  • Show basic usage with the provider
  • Demonstrate function calling if supported
  • Add streaming examples
• Add integration tests
  • Create tests/integration/mistral_integration_test.go
  • Test all provider methods
  • Add error handling tests
• Update provider integration code
  • Update pkg/util/llmutil/provider_parser.go and tests
  • Update pkg/util/llmutil/llmutil.go and tests
  • Update pkg/util/llmutil/env_vars.go and tests
  • Update pkg/util/llmutil/option_factories.go and tests
  • Update pkg/llm/provider/errors.go for Mistral errors
  • Update cmd/cli.go and cmd/config.go
• Update documentation
  • Add Mistral section to docs/user-guide/providers.md
  • Document available models and features
  • Update main README.md

0.3.3.5: AWS Bedrock provider (https://aws.amazon.com/bedrock/) 🔄 DEFERRED TO v0.4.x

• Research AWS Bedrock API and update this todo.md list
  • Investigate AWS authentication (IAM roles, access keys)
  • Study Bedrock's unified API for multiple models
  • Document available models (Claude, Llama 2, Jurassic, etc.)
  • Check streaming and function calling support
  • Identify region availability and restrictions
  • Determine model listing capabilities
• Add dedicated provider implementation
  • Create pkg/llm/provider/bedrock.go
  • Integrate AWS SDK for Go v2
  • Handle AWS authentication and region selection
  • Support multiple model families through unified API
  • Write unit tests in pkg/llm/provider/bedrock_test.go
• Add model discovery/listing support
  • Implement fetcher in pkg/util/llmutil/modelinfo/fetchers/bedrock_fetcher.go
  • Use AWS SDK to list available models
  • Handle region-specific model availability
  • Add tests for the fetcher
• Create dedicated example in cmd/examples/provider-bedrock/
  • Show AWS authentication setup
  • Demonstrate usage with different model families
  • Include region configuration
  • Show streaming if supported
• Add integration tests
  • Create tests/integration/bedrock_integration_test.go
  • Test with different AWS authentication methods
  • Test multiple model families
  • Test region-specific functionality
• Update provider integration code
  • Update pkg/util/llmutil/provider_parser.go and tests
  • Update pkg/util/llmutil/llmutil.go and tests
  • Update pkg/util/llmutil/env_vars.go and tests
  • Update pkg/util/llmutil/option_factories.go and tests
  • Update pkg/llm/provider/errors.go for AWS errors
  • Update cmd/cli.go and cmd/config.go
• Update documentation
  • Add Bedrock section to docs/user-guide/providers.md
  • Document AWS authentication setup
  • List supported models and regions
  • Include IAM permission requirements

0.3.3.6: Azure AI provider (https://azure.microsoft.com/en-us/products/ai-services/) 🔄 DEFERRED TO v0.4.x

• Research Azure AI/OpenAI Service and update this todo.md list
  • Investigate Azure authentication (API keys, Azure AD)
  • Study Azure OpenAI Service API differences
  • Document deployment model vs standard model names
  • Check for Azure-specific features
  • Identify endpoint format and regions
  • Determine model/deployment listing capabilities
• Add dedicated provider implementation
  • Create pkg/llm/provider/azure.go
  • Handle Azure-specific endpoint format
  • Support deployment names vs model names
  • Add Azure AD authentication support
  • Write unit tests in pkg/llm/provider/azure_test.go
• Add deployment discovery/listing support (if available)
  • Implement fetcher in pkg/util/llmutil/modelinfo/fetchers/azure_fetcher.go
  • Handle Azure authentication for discovery
  • Map deployments to model capabilities
  • Add tests for the fetcher
• Create dedicated example in cmd/examples/provider-azure/
  • Show API key and Azure AD authentication
  • Demonstrate deployment configuration
  • Include endpoint customization
  • Show Azure-specific features
• Add integration tests
  • Create tests/integration/azure_integration_test.go
  • Test both authentication methods
  • Test deployment name handling
  • Test region-specific endpoints
• Update provider integration code
  • Update pkg/util/llmutil/provider_parser.go and tests
  • Update pkg/util/llmutil/llmutil.go and tests
  • Update pkg/util/llmutil/env_vars.go and tests
  • Update pkg/util/llmutil/option_factories.go and tests
  • Update pkg/llm/provider/errors.go for Azure errors
  • Update cmd/cli.go and cmd/config.go
• Update documentation
  • Add Azure section to docs/user-guide/providers.md
  • Document authentication options
  • Explain deployment vs model concepts
  • Include endpoint configuration

v0.3.4: Enhanced Tool Capabilities

0.3.4.1 Runtime Tool Discovery ✅ COMPLETED (June 13, 2025)

0.3.4.5: Web API Client Advanced Features 🔄 DEFERRED TO v0.4.x

• Auto-Pagination: Automatically follow pagination links
• Rate Limiting: Respect rate limit headers with intelligent backoff
• Response Caching: Cache responses with configurable TTL
• Request Templates: Store and reuse common request patterns
• Response Transformation: Extract data using JSONPath or JQ-like queries
• Error Recovery: Smart retries with exponential backoff
• Mock Mode: Optional mock responses for testing
• Streaming Responses: Handle large response streaming
• Request/Response Middleware: Plugin system for custom processing
• Multi-tenancy: Support multiple API configurations
• Request Batching: Batch multiple requests for efficiency

0.3.4.6: Authentication System Improvements 🔄 DEFERRED TO v0.4.x

• Create AuthProvider interface with Name(), CanHandle(), Configure() methods
• Implement AuthRegistry for managing multiple providers
• Add configuration file support for custom auth mappings (YAML/JSON)
• Support provider patterns:
  • URL pattern matching (regex support)
  • Response-based detection (401 + WWW-Authenticate header)
  • OpenAPI security scheme integration
  • OAuth2 discovery via .well-known endpoints
• Create default providers for common services:
  • GitHub (including Enterprise)
  • GitLab (including self-hosted)
  • Generic Bearer token
  • Generic API key
  • Basic auth
• Ensure backward compatibility with existing detectURLSpecificAuth
• Add examples and documentation

v0.3.5: Scripting Engine Integration Support ✅ COMPLETED (June 15, 2025)

v0.3.6: Documentation Improvements & Downstream Requests

0.3.6.1: Basic Documentation Tasks ✅ COMPLETED (January 23, 2025)

0.3.6.6: Missing doc.go Files and Documentation Fixes ✅ COMPLETED (December 21, 2025)

0.3.6.2: Man Page Generation System (Priority: High)

• Implement structured man page data model
  • Create pkg/util/docs/manpage/manpage.go with core types
  • Define ManPage struct with standard sections (Name, Synopsis, Description, Options, etc.)
  • Support man page sections 1-8
• Add troff format generation
  • Create pkg/util/docs/manpage/troff.go for troff output
  • Implement proper troff macros and formatting
  • Support bold, italic, and other formatting
• Command metadata extraction
  • Create pkg/util/docs/manpage/command_extractor.go
  • Extract from Kong-based CLI structures
  • Support sub-command documentation
• Format conversions
  • Create pkg/util/docs/manpage/formatter.go
  • Support HTML output
  • Support plain text output
  • Support Markdown output
• Integration and examples
  • Add examples and cross-references support
  • Create example usage in cmd/examples/
  • Write comprehensive tests
  • Document usage in user guide

0.3.6.3: Script Documentation Extensions (Priority: High)

• Extend Documentable interface for scripts
  • Create pkg/docs/script_documentable.go
  • Add ScriptDocumentable interface extending Documentable
  • Support script language, source, parameters, and examples
• Language-specific documentation extraction
  • Create pkg/docs/script_extractor.go
  • Support Lua documentation extraction
  • Support JavaScript documentation extraction
  • Support Tengo documentation extraction
  • Create pkg/docs/language_analyzer.go for language detection
• Script parameter and schema support
  • Extract input/output parameters from scripts
  • Support JSON schema for script parameters
  • Validate parameter documentation completeness
• Script example validation
  • Create pkg/docs/example_validator.go
  • Validate script examples can execute
  • Check example inputs/outputs match schema
• Integration with existing docs system
  • Ensure scripts can use all existing doc generators
  • Support OpenAPI, Markdown, JSON export for scripts
  • Write tests and examples

0.3.6.4: Enhanced Documentation Integration (Priority: Medium)

• Create unified documentation pipeline
  • Design comprehensive documentation interfaces
  • Support tools (existing), scripts, CLI commands, APIs
  • Create consistent documentation format across all types
• Cross-reference support
  • Enable documentation to reference between tools/scripts/commands
  • Build documentation index for searching
  • Support hyperlinks in generated docs
• Export format unification
  • Ensure all documentation types can export to same formats
  • Create unified configuration for documentation generation
  • Support batch documentation generation
• Migration support for downstream projects
  • Create migration guide for go-llmspell
  • Ensure backward compatibility
  • Provide examples of bridge implementations

0.3.6.5: Documentation Utilities Refactor

• Consolidate existing doc utilities
  • Move godoc utilities to pkg/util/docs/godoc/
  • Integrate manpage utilities from 0.3.6.1
  • Create shared documentation interfaces
• Create documentation registry
  • Central registry for all documentable items
  • Support dynamic registration
  • Enable documentation discovery

0.3.6.6: Package Documentation Review ✅ COMPLETED (June 21, 2025)

0.3.6.7: Comprehensive Documentation Structure Update (Priority: High) ✅ COMPLETED (January 23, 2025)

v0.3.7: [Reserved for future features]

v0.3.8: Advanced Agent Features

0.3.8.1: State Management

• State persistence and serialization design
• Implement state storage backends (file, database)
• Add state versioning and migration
• Create examples with persistent agents

0.3.8.2: Agent Discovery

• Agent discovery and registry design
• Implement agent metadata and search
• Add dynamic agent loading
• Create agent marketplace example

0.3.8.3: Advanced Features

• Advanced merge strategies for parallel agents
• Streaming support for long-running agents
• Agent composition patterns
• Agent lifecycle management

v0.3.9: Model Context Protocol (MCP) Support

0.3.9.1: MCP Client Support

• Research MCP specification and requirements
• Design MCP client interface for agents
• Implement MCP client in pkg/agent/mcp/client
• Add MCP tool discovery and registration
• Create examples demonstrating MCP client usage
• Write comprehensive tests
• Document MCP client usage in user guide

0.3.9.2: MCP Server Support

• Design MCP server interface for exposing agents/workflows
• Implement MCP server in pkg/agent/mcp/server
• Add agent/workflow registration to MCP server
• Create example MCP server implementations
• Write comprehensive tests
• Document MCP server setup and configuration

v0.3.10: Built-in Agents Library

0.3.10.1: Text Processing Agents

• TextSummarize - intelligent summarization using LLM
• TextExtract - extract structured data from text
• TextAnalyze - sentiment, entities, keywords
• TextTranslate - language translation using LLM

0.3.10.2: Research Agents

• WebResearcher - web research with source tracking
• DocumentAnalyzer - analyze documents and PDFs
• FactChecker - verify claims against sources

0.3.10.3: Coding Agents

• CodeReviewer - review code for issues
• TestGenerator - generate tests from code
• DocWriter - generate documentation

0.3.10.4: Data Agents

• DataAnalyst - analyze datasets and generate insights
• ReportGenerator - create formatted reports
• DataCleaner - clean and validate data

0.3.10.5: Feed Agents

• NewsMonitor - monitor news feeds for keywords and topics using LLM
• FeedAggregator - aggregate and deduplicate content from multiple feeds
• FeedSummarizer - summarize feed content using LLM
• ContentCurator - curate and categorize feed content using LLM

v0.3.11: Built-in Workflow Patterns

0.3.11.1: Core Workflow Patterns

• Pipeline - sequential processing workflow
• MapReduce - parallel processing with aggregation
• Consensus - multi-agent agreement pattern
• Retry - with exponential backoff

0.3.11.2: Example Workflows

• ResearchWorkflow - research → verify → summarize → report
• CodeReviewWorkflow - analyze → review → suggest → document
• DataPipeline - ingest → clean → analyze → visualize

v0.3.12: Performance Optimization

0.3.12.1: Profiling Infrastructure (REVISIT)

• Create benchmark harness for A/B testing optimizations
• Implement visualization for memory allocation patterns
• Create real-world test scenarios for end-to-end performance

0.3.12.2: Advanced Optimizations (REVISIT)

• Implement adaptive channel buffer sizing based on usage patterns
• Add pool prewarming for high-throughput scenarios
• Reduce redundant property iterations in schema processing
• Implement more granular locking in cached objects
• Optimize zero-initialization patterns for pooled objects
• Introduce buffer pooling for string builders

0.3.12.3: Performance Validation (REVISIT)

• Document performance improvements with metrics
• Verify optimizations in high-concurrency scenarios
• Create benchmark comparison charts for before/after
• Implement regression testing to prevent performance degradation
• Add performance acceptance criteria to CI pipeline

v0.3.13: Final Polish and Stable Release

0.3.13.1: Documentation Polish

• Fix identified cross-link issues (path inconsistencies, broken links)
• Perform final consistency check across all documentation
• Update all examples to showcase v0.3.x features
• Create migration guide from earlier versions

0.3.13.2: API Refinement

• API refinement based on usage feedback
• Deprecate old patterns with migration paths
• Ensure backward compatibility where possible
• Final review and preparation for v0.4.0 stable release

Notes

All completed v0.3.x tasks have been moved to TODO-DONE.md