Roadmap

A modular, extensible architecture with clear boundaries between repository discovery, analysis, issue management and operations.

Registering and managing repositories, including their local paths and operational state.

Configurable repository scanning with progress tracking and persisted scan records.

Content fingerprinting and manifest-based incremental scanning for efficient change detection.

An extensible analyser and analysis-engine contract for introducing new languages and review types.

A Roslyn-based C# analyser maps code symbols into a structured model. C# is the initial language focus.

AI-assisted review engines apply configured prompts to repository contents via OpenAI-compatible APIs.

Findings are persisted with category, severity and status, and reconciled against the current state of the code.

Browse and inspect findings through the web interface.

Inspect the reasoning and context behind each finding.

Findings are classified and presented with severity and priority, with configurable severity rules.

A separate validation workflow re-checks findings against repository context.

Review and validation responsibilities are separated into distinct workflows.

Background workers drive validation, duplicate detection and reconciliation.

Explicit human confirmation of resolved findings is being considered for a future phase.

An operational dashboard with live status of repositories, findings and processing.

Health checks, readiness reporting and uptime monitoring.

Recovery from interrupted processing on startup with resilient background workers.

Retry handling with exponential backoff, processing gates and operational diagnostics.

Introduce new review types for additional engineering concerns.

Detect source modules that lack corresponding tests.

Generate test cases with AI assistance where a gap is identified.

Extend analysis beyond C# to additional languages.

Broaden independent validation across more finding types and contexts.