Class AnalyzeParserPerformanceTool

constructor

• new AnalyzeParserPerformanceTool(): AnalyzeParserPerformanceTool

  Returns AnalyzeParserPerformanceTool

annotations

description

name

title

execute

• execute(args: Record<string, unknown>): Promise<AnalyzeParserPerformanceOutput>

  Execute analyze-parser-performance tool to analyze parser metrics from feedback sessions

  This method loads feedback sessions from configured storage locations (project, platform, or both), extracts parser metadata from transformations, aggregates metrics across sessions, and generates prioritized improvement recommendations.

  Parameters

  • args: Record<string, unknown>

    Tool input arguments conforming to AnalyzeParserPerformanceInput interface

    • projectRoot

      Optional project root path (default: process.cwd())

    • sessionIds

      Optional specific session IDs to analyze (default: all sessions)

    • source

      Storage location strategy: 'project', 'platform', or 'both' (default: 'both')

  Returns Promise<AnalyzeParserPerformanceOutput>

  AnalyzeParserPerformanceOutput containing:

  • sessionsAnalyzed: Number of sessions that contained parser metadata
  • totalTransformations: Total transformations analyzed across all sessions
  • sourcesRead: Array of directory paths where sessions were read from
  • aggregatedMetrics: Aggregated metrics across all sessions
  • suggestions: Array of improvement suggestions sorted by priority (high → medium → low)
  • sessionMetrics: Per-session detailed metrics for comparison

  Throws

  Error if feedback storage initialization fails or critical errors occur

  Example

  // Analyze all sessions from both locations
  const result = await tool.execute({
    source: 'both'
  });
  
  // Analyze specific sessions
  const result = await tool.execute({
    sessionIds: ['session-123', 'session-456'],
    source: 'project'
  });
  Copy

  Remarks

  • Only sessions with parser metadata are analyzed
  • Sessions without parser metadata are silently skipped
  • Deduplicates sessions when reading from multiple locations
  • Aggregates metrics using weighted averages across sessions
  • Platform root can be overridden via ANGULAR_MODERNIZER_PLATFORM_ROOT environment variable
  • Suggestions are sorted by priority (high → medium → low)