Feature/blocker resolution

.github/workflows/ci.yml

@@ -39,9 +39,12 @@ jobs:
       - name: Check formatting
         run: cargo fmt --check
 
-      # TODO(REF-12): upgrade to -D warnings once clippy cleanup is complete.
       - name: Clippy
-        run: cargo clippy --all-targets
+        run: cargo clippy --all-targets -- -D warnings
 
       - name: Run tests
         run: cargo test --all
+
+      - name: Run performance tests (release build)
+        if: matrix.os == 'ubuntu-latest'
+        run: cargo test --release -- --ignored --test performance_test

CLAUDE.md

@@ -125,12 +125,12 @@ rfx query "(function_item) @fn" --ast --lang rust --glob "src/**/*.rs"
 
 ## Supported Languages
 
-**18 languages** with full symbol extraction support (functions, classes, variables, etc.):
+**15 languages** with full symbol extraction support (functions, classes, variables, etc.):
 
 - **Systems**: Rust, C, C++, Zig
 - **Backend**: Python, Go, Java, C#, PHP, Ruby, Kotlin
 - **Frontend**: TypeScript, JavaScript, Vue, Svelte
-- **Swift**: Temporarily disabled (tree-sitter version incompatibility)
+- **Swift**: Temporarily disabled — `rfx query --lang swift` emits a warning; full-text search still works but symbol queries return no results (tree-sitter-swift 0.7.x grammar incompatibility)
 
 **Symbol extraction**: Functions, classes, methods, variables (global + local), interfaces, traits, enums, attributes/annotations, and more.
 

README.md

@@ -5,7 +5,6 @@
 Reflex is a local-first, full-text code search engine. Use it from the command line, pipe it into scripts, or connect it to AI coding assistants (Claude Code, Cursor, and any MCP-compatible tool) for instant symbol lookup, dependency analysis, and codebase exploration — fully offline, fully deterministic, no cloud required.
 
 [![CI](https://github.com/reflex-search/reflex/actions/workflows/ci.yml/badge.svg)](https://github.com/reflex-search/reflex/actions/workflows/ci.yml)
-[![Tests](https://img.shields.io/badge/tests-347%20passing-brightgreen)]()
 [![License](https://img.shields.io/badge/license-MIT-blue)]()
 [![MCP Quickstart](https://img.shields.io/badge/MCP-quickstart-blue)](docs/ai-agent-integration.md)
 
@@ -85,7 +84,9 @@ When connected via MCP, your AI assistant gets these tools:
 | `count_occurrences` | Quick match statistics without full content |
 | `search_regex` | Regex pattern matching across the codebase |
 | `search_ast` | Structure-aware search via Tree-sitter AST queries |
+| `find_references` | Symbol definition + all usage sites in a single call; the primary code-navigation tool for AI agents |
 | `index_project` | Trigger or refresh the search index |
+| `check_index_status` | Check whether the index is fresh, stale, or missing; call before any search session or after git operations |
 | `get_dependencies` | All imports for a specific file |
 | `get_dependents` | All files that import a given file (reverse lookup) |
 | `get_transitive_deps` | Transitive dependency graph up to a configurable depth |
@@ -154,7 +155,7 @@ rfx index                 # Build / update the search index
 rfx index status          # Background indexing status
 rfx watch                 # Auto-reindex on file changes
 rfx stats                 # Index statistics
-rfx pulse digest          # Codebase change digest
+rfx pulse changelog       # Codebase change digest
 rfx pulse wiki            # Per-module documentation
 rfx pulse map             # Architecture diagram (Mermaid / D2)
 rfx serve --port 7878     # Local HTTP API server
@@ -190,6 +191,8 @@ Full symbol extraction (functions, classes, methods, types, etc.) for 15 languag
 **Backend:** Python, Go, Java, C#, PHP, Ruby, Kotlin  
 **Frontend:** TypeScript, JavaScript, Vue, Svelte
 
+> **Swift** is temporarily disabled (tree-sitter-swift 0.7.x grammar incompatibility). `rfx query --lang swift` emits a warning; full-text search still works.
+
 Full-text search works on **all file types** regardless of parser support.
 
 ---

benches/trigram_bench.rs

@@ -11,7 +11,6 @@ use reflex::trigram::{
 };
 use reflex::{CacheManager, Indexer, QueryEngine, QueryFilter};
 use std::fs;
-use std::path::PathBuf;
 use tempfile::TempDir;
 
 // ─── helpers ──────────────────────────────────────────────────────────────────

src/background_indexer.rs

@@ -432,12 +432,12 @@ impl BackgroundIndexer {
             });
 
             // Write batch to cache (sequential - SQLite limitation)
-            if !parsed_results.is_empty() {
-                if let Err(e) = symbol_cache.batch_set(&parsed_results) {
-                    log::error!("Failed to write symbol batch: {}", e);
-                    let mut status = status_mutex.lock().unwrap();
-                    status.2 += parsed_results.len();
-                }
+            if !parsed_results.is_empty()
+                && let Err(e) = symbol_cache.batch_set(&parsed_results)
+            {
+                log::error!("Failed to write symbol batch: {}", e);
+                let mut status = status_mutex.lock().unwrap();
+                status.2 += parsed_results.len();
             }
 
             // Update status counters
@@ -451,10 +451,10 @@ impl BackgroundIndexer {
             }
 
             // Write status every batch
-            if processed % 500 < batch_size {
-                if let Err(e) = self.write_status() {
-                    log::warn!("Failed to write status: {}", e);
-                }
+            if processed % 500 < batch_size
+                && let Err(e) = self.write_status()
+            {
+                log::warn!("Failed to write status: {}", e);
             }
         }
 

src/cache.rs

@@ -548,10 +548,10 @@ provider = "openrouter"  # Options: openai, anthropic, openrouter
             }
         }
 
-        if let Some(perf) = toml_val.get("performance") {
-            if let Some(threads) = perf.get("parallel_threads").and_then(|v| v.as_integer()) {
-                cfg.parallel_threads = threads as usize;
-            }
+        if let Some(perf) = toml_val.get("performance")
+            && let Some(threads) = perf.get("parallel_threads").and_then(|v| v.as_integer())
+        {
+            cfg.parallel_threads = threads as usize;
         }
 
         log::debug!("Loaded IndexConfig from config.toml: {:?}", cfg);
@@ -2175,7 +2175,7 @@ mod tests {
         assert_eq!(info.branch, "main");
         assert_eq!(info.commit_sha, "commit123");
         assert_eq!(info.file_count, 10);
-        assert_eq!(info.is_dirty, false);
+        assert!(!info.is_dirty);
     }
 
     #[test]
@@ -2189,7 +2189,7 @@ mod tests {
             .unwrap();
 
         let info = cache.get_branch_info("feature").unwrap();
-        assert_eq!(info.is_dirty, true);
+        assert!(info.is_dirty);
     }
 
     #[test]

src/cli/ask.rs

@@ -5,6 +5,7 @@ use owo_colors::OwoColorize;
 use std::sync::{Arc, Mutex};
 
 /// Handle the `ask` command
+#[allow(clippy::too_many_arguments)]
 pub(super) fn handle_ask(
     question: Option<String>,
     auto_execute: bool,

src/cli/deps.rs

@@ -1,5 +1,4 @@
 use crate::cache::CacheManager;
-use crate::output;
 use anyhow::Result;
 use std::path::PathBuf;
 
@@ -283,7 +282,7 @@ pub(super) fn handle_deps(
         let dependents = deps_index.get_dependents(file_id)?;
         let paths = deps_index.get_file_paths(&dependents)?;
 
-        match format.as_ref() {
+        match format {
             "json" => {
                 // Expose only paths (file_id is an internal detail)
                 let output: Vec<_> = dependents.iter().filter_map(|id| paths.get(id)).collect();
@@ -327,7 +326,7 @@ pub(super) fn handle_deps(
             // Direct dependencies only
             let deps = deps_index.get_dependencies(file_id)?;
 
-            match format.as_ref() {
+            match format {
                 "json" => {
                     let output: Vec<_> = deps
                         .iter()
@@ -399,7 +398,7 @@ pub(super) fn handle_deps(
             let file_ids: Vec<_> = transitive.keys().copied().collect();
             let paths = deps_index.get_file_paths(&file_ids)?;
 
-            match format.as_ref() {
+            match format {
                 "json" => {
                     let output: Vec<_> = transitive
                         .iter()
@@ -427,7 +426,7 @@ pub(super) fn handle_deps(
                         std::collections::BTreeMap::new();
                     for (id, d) in &transitive {
                         if *d > 0 {
-                            by_depth.entry(*d).or_insert_with(Vec::new).push(*id);
+                            by_depth.entry(*d).or_default().push(*id);
                         }
                     }
 
@@ -469,6 +468,7 @@ pub(super) fn handle_deps(
 }
 
 /// Handle --circular flag (detect cycles)
+#[allow(clippy::too_many_arguments)]
 fn handle_deps_circular(
     deps_index: &crate::dependency::DependencyIndex,
     format: &str,
@@ -594,10 +594,10 @@ fn handle_deps_circular(
                     }
                 }
                 // Show cycle completion
-                if let Some(first_id) = cycle.first() {
-                    if let Some(path) = paths.get(first_id) {
-                        println!("  → {} (cycle completes)", path);
-                    }
+                if let Some(first_id) = cycle.first()
+                    && let Some(path) = paths.get(first_id)
+                {
+                    println!("  → {} (cycle completes)", path);
                 }
             }
             if total_count > count {
@@ -641,6 +641,7 @@ fn handle_deps_circular(
 }
 
 /// Handle --hotspots flag (most-imported files)
+#[allow(clippy::too_many_arguments)]
 fn handle_deps_hotspots(
     deps_index: &crate::dependency::DependencyIndex,
     format: &str,
@@ -660,11 +661,11 @@ fn handle_deps_hotspots(
     match sort_order {
         "asc" => {
             // Ascending: least imports first
-            all_hotspots.sort_by(|a, b| a.1.cmp(&b.1));
+            all_hotspots.sort_by_key(|a| a.1);
         }
         "desc" => {
             // Descending: most imports first (default)
-            all_hotspots.sort_by(|a, b| b.1.cmp(&a.1));
+            all_hotspots.sort_by_key(|a| std::cmp::Reverse(a.1));
         }
         _ => {
             anyhow::bail!("Invalid sort order '{}'. Supported: asc, desc", sort_order);
@@ -927,6 +928,7 @@ fn handle_deps_unused(
 }
 
 /// Handle --islands flag (disconnected components)
+#[allow(clippy::too_many_arguments)]
 fn handle_deps_islands(
     deps_index: &crate::dependency::DependencyIndex,
     format: &str,
@@ -944,7 +946,7 @@ fn handle_deps_islands(
 
     // Get total file count from the cache for percentage calculation
     let cache = deps_index.get_cache();
-    let total_files = cache.stats()?.total_files as usize;
+    let total_files = cache.stats()?.total_files;
 
     // Calculate max_island_size default: min of 500 or 50% of total files
     let max_size = max_island_size.unwrap_or_else(|| {

src/cli/misc.rs

@@ -246,18 +246,17 @@ pub(super) fn handle_list_files(
         .into_iter()
         .filter(|f| {
             // Language filter: compare lowercase language name
-            if let Some(ref wanted_lang) = lang_name_filter {
-                if !f.language.to_lowercase().starts_with(wanted_lang.as_str())
-                    && f.language.to_lowercase() != *wanted_lang
-                {
-                    return false;
-                }
+            if let Some(ref wanted_lang) = lang_name_filter
+                && !f.language.to_lowercase().starts_with(wanted_lang.as_str())
+                && f.language.to_lowercase() != *wanted_lang
+            {
+                return false;
             }
             // Glob filter
-            if let Some(ref gs) = glob_set {
-                if !gs.is_match(&f.path) {
-                    return false;
-                }
+            if let Some(ref gs) = glob_set
+                && !gs.is_match(&f.path)
+            {
+                return false;
             }
             true
         })
@@ -292,6 +291,7 @@ pub(super) fn handle_mcp() -> Result<()> {
 }
 
 /// Handle the `context` command
+#[allow(clippy::too_many_arguments)]
 pub(super) fn handle_context(
     structure: bool,
     path: Option<String>,

src/cli/pulse.rs

@@ -187,6 +187,7 @@ pub(super) fn handle_pulse_map(
     Ok(())
 }
 
+#[allow(clippy::too_many_arguments)]
 pub(super) fn handle_pulse_generate(
     output: PathBuf,
     base_url: String,
@@ -376,20 +377,18 @@ pub(super) fn handle_pulse_onboard(no_llm: bool, json: bool) -> Result<()> {
     )?;
     let mut data = crate::pulse::onboard::generate_onboard_structural(&cache, modules.len())?;
 
-    if !no_llm {
-        if let Ok(provider) = crate::pulse::narrate::create_pulse_provider() {
-            let llm_cache = crate::pulse::llm_cache::LlmCache::new(cache.path());
-            let ctx = crate::pulse::onboard::build_onboard_context(&data);
-            let narration = crate::pulse::narrate::narrate_section(
-                &*provider,
-                crate::pulse::narrate::onboard_system_prompt(),
-                &ctx,
-                &llm_cache,
-                "standalone",
-                "onboard-guide",
-            );
-            data.narration = narration;
-        }
+    if !no_llm && let Ok(provider) = crate::pulse::narrate::create_pulse_provider() {
+        let llm_cache = crate::pulse::llm_cache::LlmCache::new(cache.path());
+        let ctx = crate::pulse::onboard::build_onboard_context(&data);
+        let narration = crate::pulse::narrate::narrate_section(
+            &*provider,
+            crate::pulse::narrate::onboard_system_prompt(),
+            &ctx,
+            &llm_cache,
+            "standalone",
+            "onboard-guide",
+        );
+        data.narration = narration;
     }
 
     if json {

src/cli/query.rs

@@ -2,7 +2,6 @@ use crate::cache::CacheManager;
 use crate::models::Language;
 use crate::query::{QueryEngine, QueryFilter};
 use anyhow::Result;
-use indicatif::{ProgressBar, ProgressStyle};
 use owo_colors::OwoColorize;
 use std::time::Instant;
 
@@ -28,6 +27,7 @@ pub fn truncate_preview(preview: &str, max_length: usize) -> String {
 }
 
 /// Handle the `query` subcommand
+#[allow(clippy::too_many_arguments)]
 pub(super) fn handle_query(
     pattern: String,
     symbols_flag: bool,
@@ -78,6 +78,25 @@ pub(super) fn handle_query(
         None
     };
 
+    // Warn when Swift is requested — symbol queries will return no results
+    if language == Some(Language::Swift) {
+        eprintln!(
+            "{}: Swift symbol extraction is temporarily disabled (tree-sitter-swift 0.7.x grammar incompatibility). \
+Full-text search will still work, but --symbols queries will return no results.",
+            "Warning".yellow().bold()
+        );
+    }
+
+    // Warn when --dependencies is used with a non-Rust language filter (REF-171)
+    if include_dependencies && matches!(language, Some(l) if l != Language::Rust) {
+        eprintln!(
+            "{}: --dependencies is currently only supported for Rust files. \
+No dependency data will be included for {} files.",
+            "Warning".yellow().bold(),
+            lang.as_deref().unwrap_or("non-Rust")
+        );
+    }
+
     // Parse and validate symbol kind — error on unrecognised values (REF-60)
     let kind = if let Some(s) = kind_str.as_deref() {
         let capitalized = {
@@ -123,12 +142,8 @@ pub(super) fn handle_query(
     // 3. If --paths is set and user didn't specify --limit: no limit (None)
     // 4. If user specified --limit: use that value
     // 5. Otherwise: use default limit of 100
-    let final_limit = if count_only {
-        None // --count always shows total count, no pagination
-    } else if all {
-        None // --all means no limit
-    } else if paths_only && limit.is_none() {
-        None // --paths without explicit --limit means no limit
+    let final_limit = if count_only || all || (paths_only && limit.is_none()) {
+        None // --count, --all, and --paths (without explicit --limit) all remove the result limit
     } else if let Some(user_limit) = limit {
         Some(user_limit) // Use user-specified limit
     } else {
@@ -537,7 +552,7 @@ pub(super) fn handle_query(
                                         (&content_reader_opt, file_id_for_context)
                                     {
                                         reader
-                                            .get_context_by_line(fid as u32, r.span.start_line, 3)
+                                            .get_context_by_line(fid, r.span.start_line, 3)
                                             .unwrap_or_else(|_| (vec![], vec![]))
                                     } else {
                                         (vec![], vec![])