Struct Analyzer 

pub struct Analyzer { /* private fields */ }

Implementations

impl Analyzer

pub fn new<P: AsRef<Path>>(path: P) -> Result<Self>

Create analyzer from a specific path (file or directory)

Errors

Returns an error if the path doesn’t exist or cannot be read

pub fn from_default_location() -> Result<Self>

Create analyzer from default Claude session location

Errors

Returns an error if the default Claude directory doesn’t exist

pub fn with_config(self, config: AnalyzerConfig) -> Self

Set custom configuration Used in integration tests

pub fn analyze(&self, target_file: Option<&str>) -> Result<Vec<SessionAnalysis>>

Analyze all sessions or filter by target file

Errors

Returns an error if session parsing or analysis fails

pub fn get_summary_stats(&self) -> Result<SummaryStats>

Get summary statistics across all sessions

Errors

Returns an error if analysis fails

pub fn calculate_agent_tool_correlations( &self, tool_invocations: &[ToolInvocation], ) -> Vec<AgentToolCorrelation>

Calculate which agents use which tools Returns correlations sorted by usage count (descending)

pub fn calculate_tool_statistics( &self, tool_invocations: &[ToolInvocation], ) -> IndexMap<String, ToolStatistics>

Calculate comprehensive tool statistics

pub fn calculate_category_breakdown( &self, tool_invocations: &[ToolInvocation], ) -> IndexMap<ToolCategory, u32>

Calculate category breakdown

pub fn detect_tool_chains( &self, tool_invocations: &[ToolInvocation], ) -> Vec<ToolChain>

Detect sequential tool usage patterns (tool chains)

Arguments

• tool_invocations - List of tool invocations to analyze

Returns

A vector of ToolChain instances representing detected patterns

Algorithm

1. Group tools by session
2. Sort by timestamp within each session
3. Use sliding windows (2-5 tools) to find sequences
4. Group identical sequences across sessions
5. Calculate frequency, timing, and success rate
6. Filter chains that appear at least twice