Struct ToolRunner 

pub struct ToolRunner { /* private fields */ }

Executes diagnostic tools and captures their output.

Each tool is run as a subprocess with configurable timeout. Output is captured and fed through the parser identified by ToolConfig::parser.

Implementations

impl ToolRunner

pub fn new(timeout_secs: u64) -> Self

Create a new ToolRunner with the given per-tool timeout in seconds.

pub fn run_tool( &self, tool: &ToolConfig, project_path: &Path, ) -> (ToolResult, Vec<L1Finding>)

Run a single tool and parse its output.

Contract

• Binary not found: ToolResult { success: false, error: "spawn" message }
• Tool timeout: kill child, ToolResult { success: false, error: "Timeout" }
• Tool crashes (non-zero exit) with parseable output: success: true (linters exit non-zero when findings exist)
• Tool crashes with unparseable output: success: false
• Parse error: ToolResult { success: false, error: "Parse error: ..." }
• After parsing, injects tool.name into each L1Finding.tool [PM-6]

pub fn run_tools_parallel( &self, tools: &[&ToolConfig], project_path: &Path, ) -> (Vec<ToolResult>, Vec<L1Finding>)

Run multiple tools in parallel, collecting results.

Contract

• One tool failure does not block others
• Results are in deterministic order (same as input tools order)
• All findings have tool name injected [PM-6]
• Single tool or empty list: runs sequentially (no thread overhead)