ios-mcp-code-quality-server

Exposes iOS code quality analysis tools through the Model Context Protocol, allowing Claude and other MCP-compatible clients to invoke linting, complexity analysis, and code smell detection on Swift/Objective-C codebases. Implements MCP server architecture with tool registration endpoints that map to underlying analysis engines, enabling LLM-driven code review workflows without direct filesystem access from the client.

Implements MCP server-side tool registry that maps high-level analysis requests (e.g., 'analyze file for violations') to underlying iOS code quality tools via standardized schema definitions. Uses MCP's tool definition protocol to expose analysis capabilities with typed parameters, return schemas, and error handling, allowing clients to discover and invoke tools with IDE-like autocomplete and validation.

Parses analysis tool output (SwiftLint JSON, etc.) and enriches violation reports with source code context, including surrounding lines, violation location metadata, and rule documentation. Returns structured violation objects that include file path, line/column numbers, violation severity, rule identifier, and suggested fixes, enabling LLMs to reason about violations in their source context without requiring separate file reads.

Orchestrates analysis across multiple Swift/Objective-C files in an iOS project, aggregating results into project-level metrics (total violations, violation distribution by rule, severity breakdown, code quality score). Implements batching and parallel execution where possible to reduce total analysis time, and returns both file-level details and project-level summaries in a single response.

Exposes configuration endpoints that allow clients to customize which iOS code quality rules are enabled, set severity levels (error/warning/info), and define project-specific rule exceptions. Implements configuration persistence (via config files or server state) and applies custom rules to all subsequent analysis invocations, enabling teams to tailor analysis to their coding standards without modifying the analysis tools directly.

Analyzes detected violations and generates actionable remediation suggestions by combining rule documentation, violation context, and code patterns. For common violations (unused variables, naming conventions, etc.), provides code snippets showing the corrected version. Leverages rule metadata and heuristics to suggest fixes without requiring external LLM calls, enabling fast, deterministic remediation guidance.

Classifies code quality violations by severity level (error, warning, info) based on analyzer output and configured rules, enabling clients to prioritize remediation. Implements severity mapping from tool-specific violation types to standard severity levels, and supports custom severity overrides based on project configuration or violation patterns.

Generates automated fix suggestions for detected violations where analyzers provide remediation hints, and implements code transformation logic to apply fixes to source files. Parses analyzer-provided fix information (line/column ranges, replacement text) and generates structured fix objects that clients can apply or present to users for approval.

Supports analysis at multiple scopes (single file, directory, entire project) with appropriate context and performance optimizations for each scope. Implements scope detection logic to determine analysis boundaries, manages file discovery for directory/project scopes, and optimizes analyzer invocation based on scope (e.g., single-file analysis skips project-wide checks).

Monitors the health and availability of registered iOS analyzers, detects and reports analyzer failures, and implements graceful degradation when tools are unavailable. Implements health checks to verify analyzer installation and accessibility, captures analyzer error output, and returns structured error information to clients.