cqs 1.25.0

Code intelligence and RAG for AI agents. Semantic search, call graphs, impact analysis, type dependencies, and smart context assembly — in single tool calls. 54 languages + L5X/L5K PLC exports, 91.2% Recall@1 (BGE-large), 0.951 MRR (296 queries). Local ML, GPU-accelerated.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22

//! Train-data command — generate training data from indexed code

use anyhow::Result;

use cqs::train_data::TrainDataConfig;

pub fn cmd_train_data(config: TrainDataConfig) -> Result<()> {
    let _span = tracing::info_span!("cmd_train_data").entered();
    let stats = cqs::train_data::generate_training_data(&config).map_err(|e| anyhow::anyhow!(e))?;

    println!(
        "Generated {} triplets from {} repos ({} commits processed, {} skipped)",
        stats.total_triplets, stats.repos_processed, stats.commits_processed, stats.commits_skipped
    );
    if stats.parse_failures > 0 {
        println!("  {} parse failures", stats.parse_failures);
    }
    for (lang, count) in &stats.language_counts {
        println!("  {}: {} triplets", lang, count);
    }
    Ok(())
}